Reliability of Expert Evidence in International Disputes

Michigan Journal of International Law
Volume 38 | Issue 2
2017
Reliability of Expert Evidence in International

Disputes
Matthew W. Swinehart
United States Department of the Treasury
Follow this and additional works at: http://repository.law.umich.edu/mjil

Part of the Courts Commons, Evidence Commons, International Law Commons, and the
Organizations Law Commons
Recommended Citation
Matthew W. Swinehart, Reliability of Expert Evidence in International Disputes, 38 Mich. J. Int'l L. 287 (2017).
Available at: http://repository.law.umich.edu/mjil/vol38/iss2/6
This Article is brought to you for free and open access by the Journals at University of Michigan Law School Scholarship Repository. It has been
accepted for inclusion in Michigan Journal of International Law by an authorized editor of University of Michigan Law School Scholarship Repository.
For more information, please contact [email protected].
ARTICLE
RELIABILITY OF EXPERT EVIDENCE IN

INTERNATIONAL DISPUTES
Matthew W. Swinehart*
INTRODUCTION
Even as the legal community, politicians, and the public at large con-
tinue to scrutinize the legitimacy of international dispute resolution,1 one
of the most influential features to have emerged in modern international
law—the use of expert witnesses—has gone largely overlooked. Apart
from the broader debates about whether and to what extent states should
subject policy choices to international law or how treaty partners should
design mechanisms to resolve disputes over those choices,2 stands the sig-
nificant role that expert witnesses play in those disputes. Where the dis-
puting parties appoint experts, the adjudicator often must referee a “battle
of the experts” and choose between two competing views and outcomes.
Where the adjudicator has appointed an expert, the struggle is to ensure
that the ultimate decision making authority is not delegated to the expert.
Many recent high-profile international legal disputes—including a dispute
* Matthew Swinehart is a legal advisor for trade and investment agreement

negotiations at the United States Department of the Treasury. The views expressed here are
the author’s own and not necessarily the views of the United States government or the
Department of the Treasury. This paper has benefitted from conversation with colleagues
including Christopher Bradley, Suzanne Garner, and the participants of the 2016 American
Society of International Law Biennial Conference of the International Economic Law
Interest Group at Georgetown University Law Center, at which an earlier version of this
paper was presented. Many thanks are also owed to the excellent editorial staff of the
Michigan Journal of International Law.
1. See, e.g., Zeeshan Aleem, Why Trump’s Plan to Ignore the World Trade Organiza-
tion Isn’t as Reckless as it Sounds, VOX (Mar. 8, 2017), http://www.vox.com/policy-and-poli
tics/2017/3/8/14766228/trump-trade-wto (“[T]he Trump administration’s skepticism of the
WTO isn’t new.”); Catherine A. Rogers, The Politics of International Investment Arbitration,
12 SANTA CLARA J. INT’L L. 223, 233 (2013) (noting a pervasive sense of a “legitimacy crisis”
in the investment arbitration world); James Surowiecki, The Financial Page: Trade Agreement
Troubles, THE NEW YORKER, June 22, 2015, at 26 (arguing that investor-state dispute settle-
ment is “outdated and unnecessary” and that “including them in trade agreement under-
mines the broader case for free trade, by making it look like exactly what people fear—a
system designed to put corporate interests above public ones”).
2. These debates—whether we should enshrine rules in international agreements and
subject those rules to dispute settlement—are the “substance of government policy.” JAN
PAULSSON, DENIAL OF JUSTICE IN INTERNATIONAL LAW 232 (2005). This Article does not
wade into them. But we must ensure that the international tools for implementing those
policy choices—whatever they may be—produce sound decisions and uphold the legitimacy
of the decision-making process.
287
288 Michigan Journal of International Law [Vol. 38:287
between China and the Philippines over their territorial seas and a tobacco
company’s challenge to Uruguay’s plain packaging rules for cigarettes—
have centered on technical evidence presented by experts.
This is a relatively new feature of international disputes, and we are
still in the early days of dealing with its consequences. For the first two
centuries of modern international dispute resolution, tribunals and courts
sought out expert advice only on rare occasions, usually as a last resort.
But reliance on experts has increased dramatically in the last forty years.
Human activity itself has become more complex—more scientific, more
specialized, more reliant on experts—and so too have disputes that arise
out of that activity. At the same time, international agreements have con-
tinued evolving to tackle progressively more complex and technical issues,
often through express reliance on science, economics, and other special-
ized fields.
Today, as a result, parties to international disputes routinely put for-
ward expert evidence to support their arguments, and courts and tribunals
often appoint their own experts to bridge the gap between law and other
fields. Economists and accountants opine on the amount of damages owed
to aggrieved investors; scientists evaluate the basis of environmental and
health regulations; historians and anthropologists testify on the social, po-
litical, and historical contexts of armed conflict and violence; and engi-
neers explain the development of intellectual property and technologies.
Many international tribunals and courts have failed to confront this
reality. Most do not ask whether the expert evidence before them even
meets a minimum threshold of reliability. And when they do ask, the as-
sessment routinely lacks rigor and transparency. This neglect can lead to
reliance on unreliable expert evidence, unfair rejection of evidence that is
reliable, or opaque and confused attempts to resolve conflicting expert
views. Readers of the resulting decisions are often left wondering about
the role, if any, that experts played in resolving the disputes.
The legitimacy of international legal systems—systems predicated on
the consent of sovereign nations and entrusted to pass judgment on the
public policy choices of those nations—depends on decision making that
at least appears analytically sound. International courts and tribunals
“must ensure that they are in a position to appreciate the disputed policy
choices made by the states appearing before them” through rigorous deci-
sion making, defensible analytical frameworks, and thorough appreciation
of the facts of each dispute that comes before them.3 This means that the
“instruments and processes of international law must provide means for
scientific evidence,” and other expert evidence, “to be sifted, understood,
and translated into law.”4 While decision makers do not “have to under-
3. CAROLINE E. FOSTER, SCIENCE AND THE PRECAUTIONARY PRINCIPLE IN INTERNA-

TIONAL COURTS AND TRIBUNALS: EXPERT EVIDENCE, BURDEN OF PROOF AND FINALITY 77
(2011).
4. Harlan Cohen et al., An Introduction: Confronting Complexity, in AM. SOC’Y INT’L
L., PROCEEDINGS OF THE 106TH ANNUAL MEETING 1, 1 (2013); see also FOSTER, supra note
3, at 77 (“Sound decision-making is essential in an international legal system where submis-
Winter 2017] Reliability of Expert Evidence in International Disputes 289
stand a proposition to be justified in believing it,” they do need to “be able

to repose a justified trust in the truthfulness and expertise of the person
who assures [them] that the proposition is true.”5
Importantly, we do find examples where international law has made
careful use of expert evidence, and this Article seeks to aggregate best
practices from a broad range of substantive areas, including trade and in-
vestment, criminal prosecutions, boundary disputes, the law of the sea,
post-war and peace commissions, and environmental law. The core of this
proposal is that international adjudicators should adopt an analytical
framework for determining whether an expert’s testimony is reliable, ap-
ply that framework to each expert, clearly set out in the written decision
the framework and how it was applied, and take into account the reliabil-
ity analysis when weighing the evidence presented.
The framework could take the form of a nonexhaustive checklist that
would focus on the reliability of an expert’s methodology and the applica-
tion of that methodology to the facts of a dispute. What is the relevant
question and is expert evidence even necessary to answer it? Is the ex-
pert’s methodology one that others in the field would use? Has the expert
acknowledged and remained within the methodology’s practical limita-
tions? And has the expert presented the methodology and any conclusions
in way that is useful to non-experts? Asking those questions and others
can help ensure that international law uses expert evidence in a more rig-
orous, defensible, and transparent manner.
Unlike some other efforts to refine the analytical legitimacy of inter-
national adjudication, improving the clarity and transparency with which
international arbitrators and judges evaluate expert evidence is in most
cases an improvement that can be made by the adjudicators themselves.
Existing agreements already vest adjudicators with broad authority and
discretion in evaluating expert witness testimony. Harnessing that discre-
tion with a clearly stated decision-making framework has the potential to
improve legitimacy and to lead to more correct legal outcomes.
Part I below traces the historical trends in the use of expert evidence
in international disputes, from the scattered reliance on experts in the
nineteenth and early twentieth centuries to the ubiquity of experts in mod-
ern disputes. With that perspective, Part II examines how decision makers
have attempted to ensure reliability of the expert evidence that is flooding
the evidentiary records of international disputes, while Part III outlines
the many problems that still remain. Finally, Part IV proposes a non-ex-
haustive and nonbinding checklist of questions for analyzing the reliability
of any type of expert evidence.
sion to jurisdiction is voluntary and many of the interests at stake are important, sensitive
and complex.”).
5. Richard A. Posner, The Law and Economics of the Economic Expert Witness, 13 J.
ECON. PERSP. 91, 96 (1999).
I. HISTORICAL TRENDS IN THE USE OF EXPERT EVIDENCE

Exploring the historical context is critical to understanding how inter-
national courts and tribunals can best make use of expert evidence. Al-
though commonplace today, reliance on expert evidence in international
disputes is a relatively new phenomenon, rare until the last forty years.
The first hundred years of the modern era of international dispute set-
tlement—between the 1794 Jay Treaty6 and World War I—provide few
salient examples of reliance on expert testimony. Although more than 300
tribunals operated during this time,7 and the resulting decisions have been
described as producing a “vast international arbitral jurisprudence,”8 the
century provides few salient examples of reliance on expert testimony.
The peace treaties and other international instruments signed in the
wake of the World Wars led to a modest increase in reliance on expert
testimony.9 The various Mixed Arbitral Tribunals established under the
World War I peace treaties used expert witnesses particularly frequently.10
In Caisse d’Assurances des Glaceries v. Germany, for example, the Bel-
gian-German tribunal sought expert testimony on the value of a factory
6. Arbitral or judicial settlement of international disputes has been described as “ir-

regular and spasmodic” before the 1794 signing of the Jay Treaty between Great Britain and
the United States. See JACKSON RALSTON, INTERNATIONAL ARBITRATION FROM ATHENS TO
LOCARNO 191 (1929). Because that settlement “assumed a certain regularity and system”
after the Jay Treaty, the Treaty is generally considered the beginning of the modern era of
international dispute settlement. Id. at vii, 191.
7. See RALSTON, supra note 6, at 345–54 (listing 304 “arbitral and other judicial tribu-
nals functioning between nations” from 1794 to before the World War I tribunals that began
in 1919); see also EVANS DARBY, INTERNATIONAL TRIBUNALS: A COLLECTION OF THE VARI-
OUS SCHEMES WHICH HAVE BEEN PROPOUNDED, AND OF INSTANCES SINCE 1815 285–304
(1899) (listing 158 international tribunals between 1815 and 1899); Paulsson, supra note 2, at
131 n.1 (noting that some of these tribunals resolved multiple disputes).
8. PAULSSON, supra note 2, at 131.
9. See GILLIAN M. WHITE, THE USE OF EXPERTS BY INTERNATIONAL TRIBUNALS
80–81 (1965) (noting that treaties and rules of procedure increasingly “ma[d]e use of expert
help in the form of testimony, reports and investigations on the spot” and that “several of the
tribunals established after the world wars . . . have availed themselves of such powers”).
Several World War I peace treaties provided for the use of experts. See id. at 50 (citing as
examples Article 56 of the Rules of the Belgian-German tribunal, Articles 57 and 60 of the
Rules of the Franco-German tribunal, Article 3 of the Agreement of June 30, 1921 Between
the United States of America and Norway for the Submission to Arbitration of Certain
Claims of Norwegian Subjects, and Article 3 of the Arbitration Protocol between France and
Haiti of September 10, 1913).
10. See DURWARD V. SANDIFER, EVIDENCE BEFORE INTERNATIONAL TRIBUNALS 325
n.137 (1975) (citing Huret v. État Allemand (Franco-German Tribunal), Vol. 1, RECUEIL DES
DÉCISIONS DES TRIBUNAUX ARBITRAUX MIXTES 98, 488 (1921); Lirens v. État Allemand
(German-Belgian Tribunal), 2 RECUEIL DES DÉCISIONS DES TRIBUNAUX ARBITRAUX
MIXTES 82; Companie d’Electricité de Sofia et de Bulgarie v. État Bulgare et Municipalite de
Sofia (Bulgarian-Belgian Tribunal), 3 RECUEIL DES DÉCISIONS DES TRIBUNAUX ARBITRAUX
MIXTES 308, 323 (1923)). Many of the experts involved in these Tribunals were engaged to
assist in valuations. For example, the Franco-German tribunal in Huret v. State of Germany
appointed an expert to value a specific make and model of a car that had been requisitioned.
See 1 RECUEIL DES DÉCISIONS DES TRIBUNAUX ARBITRAUX MIXTES 98, 488; see also WHITE,
supra note 9, at 51-52.
expropriated by the Germans, including the “sum required . . . to construct

[it] in Germany,” what its “current value” would have been, and “what
profits it might have earned to date.”11
Although the commissions established under the 1947 peace trea-
ties,12 the 1951 peace treaties with Japan,13 and the early 1950s Bonn-Paris
Conventions with Germany14 were broadly empowered to consider expert
evidence or to develop their own procedures,15 it does not appear that the
commissions established after World War II made widespread use of ex-
perts.16 As an example, the Arbitral Commission on Property, Rights, and
Interest in Germany, established in 1954 and described as “one of the
most active of the . . . international courts” of the mid-century period,
adjudicated more than one hundred cases but referred a matter to experts
in only one reported decision.17 In that case, Greece and the Firm Apos-
tolidis v. Germany, experts determined the going price of chrome in Ger-
many to form the basis of the Commission’s compensation calculation.18
Of the World War II commissions, the Franco-Italian Commission is the
11. 3 RECUEIL DES DÉCISIONS DES TRIBUNAUX ARBITRAUX MIXTES, supra note 10, at
261, 270; see also WHITE, supra note 9, at 51-52.
12. See, e.g., Treaty of Peace with Italy art. 83(3), Feb. 10, 1947, 61 Stat. 1245 (empow-
ering the Italian Commissions to “adopt [ ] rules conforming to justice and equity”); Rules of
Procedure of the Franco-Italian Commission art. 14, translated in WHITE, supra note 9, at 53
(“The Commission may decide . . . to arrange for the carrying out of any expert enquiry with
the co-operation of all necessary technicians, interpreters or translators.”).
13. See, e.g., Rules of Procedure of the United States-Japanese Property Commission
art. 13(E), in RICHARD B. LILLICH & GORDON A. CHRISTENSON, INTERNATIONAL CLAIMS:
THEIR PREPARATION AND PRESENTATION app. F, at 144 (1962) (“The [United States-Japa-
nese Property] Commission may appoint an expert and request him to submit an opinion in
writing.”).
14. Convention on the Settlement of Matters Arising out of the War and the Occupa-
tion, Oct. 23, 1954, [1955] 6 U.S.T. & O.I.A. 4411, T.I.A.S. No. 3425 (“The Commission shall
have the power . . . to request expert opinion.”); Rules of Procedure of the Arbitral Commis-
sion on Property, Rights, and Interest in Germany, Rule 22(a), in 1 DECISIONS OF THE ARBI-
TRAL COMMISSION ON PROPERTY, RIGHTS, AND INTEREST IN GERMANY 196 (1958)
[hereinafter Rules of Procedure of the German Arbitral Commission] (“The Commission
shall have power, upon application of a party or of its own motion . . . to request expert
opinion.”).
15. See WHITE, supra note 9, at 52–54. Although all the 1947 peace treaties “included
identical provisions for the settlement of disputes . . . only those in the peace treaty with Italy
were put into effect.” Id.
16. See WHITE, supra note 9, at 55 (noting that “none of the published decisions” of
the United States-Japanese Property Commission “mention that the Commission ever made
use of the power”). This may have been because “in no case did it prove necessary to con-
sider detailed evidence of damages.” Lionel M. Summers & Arnold Fraleigh, The United
States-Japan Property Commission, 56 AM. J. INT’L L. 407, 411 (1962).
17. WHITE, supra note 9, at 57 (citing Government of the Kingdom of Greece and the
firm Apostolidis v. Federal Republic of Germany, 3 DECISIONS OF THE ARBITRAL COMMIS-
SION ON PROPERTY, RIGHTS, AND INTERESTS IN GERMANY 329, 364 (1960)).
18. Government of the Kingdom of Greece and the firm Apostolidis v. Federal Republic
of Germany, 3 DECISIONS OF THE ARBITRAL COMMISSION ON PROPERTY, RIGHTS, AND IN-
TERESTS IN GERMANY 329, 364 (1960).
only one cited as providing “many examples of the use of independent

experts.”19
The standing dispute resolution institutions of the era also rarely re-
lied on experts. The Permanent Court of Arbitration did not rely on ex-
pert testimony for its first several decades, even though it possessed
express authority to appoint its own experts.20 In its eighteen years of exis-
tence, between 1922 and 1940, the League of Nations’s Permanent Court
of International Justice (PCIJ) sought out expert testimony in just one dis-
pute, the widely cited Factory at Chorzów case.21 And in the post-war era,
the International Court of Justice (ICJ), the heir to the PCIJ, continued
the trend, relying on expert testimony only once, in the 1949 Corfu Chan-
nel case.22
There were exceptions. One class of cases in particular—boundary
disputes—consistently bucked the general trend.23 From the early days of
modern international dispute resolution, decision makers regularly relied
on expert evidence on geography and hydrogeology,24 including the St.
Croix River dispute between Great Britain and the United States in
1798.25 Other than questions of delineating sovereign boundaries, there
19. WHITE, supra note 9, at 53. A prominent example from the Commission is the
Ousset Claim. See 5 RECUEIL DES DÉCISIONS DE LA COMMISSION DE CONCILIATION
FRANCO-ITALIENNE 36; 22 I.L.R. 312 (1955). During the first decade of existence, neither the
European Commission of Human Rights nor the European Court of Human Rights exer-
cised its authority to rely on expert testimony. See WHITE, supra note 9, at 62, 66.
20. See WHITE, supra note 9, at 36, 74, 138. In one case, Lighthouses Arbitration (Fr. v.
Greece), the Permanent Court of Arbitration ordered expert testimony, but the parties set-
tled the claims before experts were appointed. 12 R.I.A.A. 155; 23 I.L.R. 659 (1956).
21. Factory at Chorzów (Ger. v. Pol.) (Indemnity), 1928 P.C.I.J. (ser. A) No. 17, at
136-51); see also WHITE, supra note 9, at 40.
22. See Corfu Channel (U.K. v. Alb.), Merits, 1949 I.C.J. Rep. 4 (Apr. 9) (relying on
tribunal-appointed experts to determine liability); Corfu Channel (U.K. v. Alb.), Judgment,
1949 I.C.J. Rep. 244 (Dec. 15) (relying on tribunal-appointed experts to determine the
amount of compensation owed).
23. Adjudicators under Roman law relied on experts in land disputes, see ARTHUR
ENGLEMANN, A HISTORY OF CONTINENTAL CIVIL PROCEDURES 361–62 (1928), as did Italian
and French courts since the middle ages, with codification of expert procedures in a 1667
French ordinance, see WHITE, supra note 9, at 16.
24. See, e.g., A. de LAPRADELLE & N. POLITIS, 2 RECUEIL DES ARBITRAGES INTERNA-
TIONAUX 574, 579 (2d ed. 1957) (Eastern Boundary (S. Afr. v. Orange Free State) (1870));
Manica Plateau Arbitration (Gr. Brit. v. Port.), 28 R.I.A.A. 283, 308 (1897)); Cordillera of
the Andes Boundary Case (Arg. v. Chile), 9 R.I.A.A. 29, 39–40 (Perm. Ct. Arb. 1902); L’Oeil
de la Mer (Austria v. Hung.) (1902), 8 REV. DE DROIT INT’L ET DE LÉGISLATION COMP. 162
(2d ser. 1906); 3 MARTENS NOUVEAU RECUEIL GÉNÉRALE DE TRAITÉS (ser. 71) 465; Costa
Rica-Panama Boundary (Costa Rica v. Pan.), 11 R.I.A.A. 518, 539-40 (1914); Honduras Bor-
ders (Guat. v. Hond.), 2 R.I.A.A. 1307, 1353 (Int’l Central American Trib., 1933); Gran
Chaco Dispute (Bol. v. Para.), 3 R.I.A.A. 1817,1822-25 (1938).
25. See JOHN BASSETT MOORE, HISTORY AND DIGEST OF THE INTERNATIONAL ARBI-
TRATIONS TO WHICH THE UNITED STATES HAS BEEN A PARTY 23–24, 29–30 (1898). The
expert testimony—that of land surveyors and astronomers—may have played a significant
role in the outcome, as the proceedings were twice postponed while their surveys and reports
were completed. Id. at 23–24. The award was rendered only ten days after receipt of their
final map. Id. at 24. The map was annexed to the award. Id. at 30.
were relatively few examples of reliance on expert evidence—some re-

lated to valuations, science, document authentication, and engineering, for
example—in the nineteenth century and the twentieth century through the
1970s.26
The trend in the last forty years has been very different, as tribunals
and courts have dramatically increased the frequency and significance of
their reliance on expert evidence. Many factors appear to have contrib-
uted to the trend, which parallels the increased reliance on experts in
other decision-making processes, including international governance
outside legal dispute resolution.27 The first order explanation is simply
that human activity itself has become more complex—more scientific,
more specialized, more reliant on experts—and so too have disputes that
arise out of that activity.28 But international law itself has played a role,
too.
The more robust procedural and evidentiary frameworks in interna-
tional agreements may also explain some of the rise in reliance on experts.
Express authorization to rely on expert testimony gradually became a
ubiquitous feature of procedural rules of standing tribunals and courts in
the mid-twentieth century.29 International dispute settlement is also in-
26. See, e.g., Lehigh Valley Railroad Co. (U.S. v. Ger.), 8 R.I.A.A. 104, 121 (1932)
(analyzing over one thousand pages of evidence submitted by document authentication ex-
perts); I.R. Clark (U.S. v. Mex.), 4 R.I.A.A. 415, 415 (1928) (noting that the Mexico-U.S.
Commission would give “careful consideration” to the “expert testimony” of a physician);
Marguerite de Joly de Sabla (U.S. v. Pan.), 6 R.I.A.A. 358, 362, 367 (1933) (relying in part on
the testimony of a licensed surveyor); Fur Seal Arbitration (U.S. v. Gr. Brit. 1892), MOORE,
supra note 25, at 755, 813–14 (examining the testimony of fisheries experts).
27. See Monika Ambrus et al., The Role of Experts in International and European
Decision-Making Processes: Setting the Scene, in THE ROLE OF “EXPERTS” IN INTERNA-
TIONAL AND EUROPEAN DECISION-MAKING PROCESSES: ADVISORS, DECISION MAKERS OR
IRRELEVANT ACTORS? 1, 5–6 (M. Ambrus et al. eds., 2014) (“Experts are called upon to
provide input into ever more complex decision-making processes. They are involved in pol-
icy making, law-making, and implementation and enforcement efforts.”).
28. See Anna Riddell, Scientific Evidence in the International Court of Justice:
Problems and Possibilities, in 20 FINNISH .B. INT’L L. 229, 230 (2009) (“Major technological
and scientific developments in recent decades have ensured that scientific knowledge has
developed an important role in the administration of law.”); Dushyant Dave, Contemporary
Practice in the Conduct of Proceedings: Techniques for Eliciting Expert Testimony – How
Party-Appointed Experts Can Be Made Most Useful, in INTERNATIONAL ARBITRATION 2006:
BACK TO BASICS? 813, 818 (Albert Jan van den Berg ed., 2006) (“[I]nternational tribunals
have become much more dependent on expert testimony, and the accelerating process of
scientific and technological advance enhances this tendency.”).
29. See, e.g., Consolidated Version of the Protocol on the Statute of the Court of Jus-
tice of the Coal and Steel Community, art. 25, Apr. 18, 1951 (“The Court may at any time
entrust any individual, body, authority, committee or other organisation it chooses with the
task of holding an inquiry or giving an expert opinion; to this end it may compile a list of
individuals or bodies approved as experts.”); Maarten Bos, The Franco-Italian Conciliation
Commission, 22 NORDIC J. OF INT’L L. 133, 154 (1952)
Article 14, paragraph 3, [of the Rules of Procedure of the Franco-Italian Concilia-
tion Commission] confers upon the Commission the right to have examined by
experts nominated by itself the reports of the experts nominated by the parties.
creasingly open to nongovernmental groups, international organizations,

academics, and the public at large, including through the submission of
non-disputing party submissions and amicus curiae briefs, many of which
may present expert-like opinions.30
At the same time, this is partly because international agreements over
the years have evolved to tackle ever more complex issues, themselves
becoming more technical, more specialized, and more tailored to the
problems that they seek to address.31 Countries negotiating trade agree-
ments, once concerned primarily with reducing tariff barriers and quotas,
have shifted focus to other internal laws and regulations, such as those
concerning health, safety, and the environment, that may act as unfairly
trade restricting or distorting.32 Agreements also often require or en-
Under paragraph 4 of this Article the Commission may visit the places concerned
in a dispute and may hear its own experts there. The Agents or their substitutes
are, by paragraph 5, entitled to be present.
30. See Joost Pauwelyn, The Use of Experts in WTO Dispute Settlement, 51 INT’L
COMP. L.Q. 325, 325 (2002) (“[A]n increasing number of ‘outsiders’ or amici curiae, such as
NGOs, but also industry and academics, have pressed their (expert) opinion on WTO panels
and the Appellate Body.”); OFF. OF THE U. S. TRADE REPRESENTATIVE: THE TRANS-PAC.
PARTNERSHIP art. 9.23(3), https://ustr.gov/sites/default/files/TPP-Final-Text-Investment.pdf
(last visited Feb. 5, 2017) (“After consultation with the disputing parties, the tribunal may
accept and consider written amicus curiae submissions regarding a matter of fact or law
within the scope of the dispute that may assist the tribunal in evaluating the submissions and
arguments of the disputing parties from a person or entity that is not a disputing party but
has a significant interest in the arbitral proceedings.”); NAFTA Free Trade Comm’n, State-
ment of the Free Trade Commission on Non-Disputing Party Participation, (Oct. 7, 2004)
http://.naftaclaims.com/commissionfiles/Nondisputing-en.pdf (recommending that NAFTA
investor-state tribunals adopt certain procedures regarding non-disputing party submissions).
A number of investment arbitration decisions have relied on amicus curiae submissions
presenting scientific and other technical evidence. See, e.g., Philip Morris Brands Sarl v. Uru-
guay, ICSID Case No. ARB/10/7, Award ¶¶ 394, 396 (2015) (relying on a World Health
Organization submission); Methanex Corp. v. United States, Decision of the Tribunal on
Petitions from Third Persons to Intervene as “Amici Curiae” ¶ 53 (Jan. 15, 2001) (determin-
ing that a tribunal has the authority to accept amicus written submissions under Article 15(1)
of the UNICTRAL Arbitration Rules).
31. See Kate Miles, Climate Change: Trading, Investing and the Interaction of Law,
Science and Technology, in SCIENCE AND TECHNOLOGY IN INTERNATIONAL ECONOMIC LAW:
BALANCING COMPETING INTERESTS 155, 155 (B. Mercurio & Kuei-Jung Ni eds., 2014) (“The
ever-increasing complexity of the issues with which international law is required to grapple
has recently seen, among other effects, a notable visibility of science and technology in inter-
national law.”).
32. See Lukasz Gruszczynski, Science and the Settlement of Trade Disputes in the
World Trade Organization, in SCIENCE AND TECHNOLOGY IN INTERNATIONAL ECONOMIC
LAW: BALANCING COMPETING INTERESTS 11, 24 (B. Mercurio & Kuei-Jung Ni eds., 2014)
(“GATT 1947 was predominantly concerned with tariff barriers and quotas. Its success in
reducing these types of obstacle has shifted the attention of countries to market access
problems caused by internal measures (i.e.[,] non-tariff barriers). In consequence, potential
disputes under WTO law have become not only more complex (than tariff barriers and quo-
tas), but also more politically contentious.”); Robert Howse, Democracy, Science, and Free
Trade: Risk Regulation on Trial at the World Trade Organization, 98 MICH. L. REV. 2329,
2329 (2000) (“Traditionally, free trade rules were about constraining border measures such as
tariffs and quantitative restrictions on imports. Increasingly, however, such rules include re-
quirements and constraints addressed directly to domestic regulation.”).
courage countries to adopt international standards or establish scientific

evidence requirements for certain types of technical (especially environ-
mental, health, and safety) regulation.33 International legal systems that
have emerged in the last forty years, some with inherently technical man-
dates and legal norms that are inextricably bound up with questions of
specialized knowledge, routinely elicit and rely on expert testimony.34 Re-
cent agreements also contain procedural provisions that contemplate tri-
bunal-appointed experts on specific technical issues35 or that otherwise
encourage reliance on expert evidence.36 Partly as a result of these provi-
sions, international disputes increasingly arise not only with respect to past
injury but also risks of future harm—to human health or the environment,
for example—and expert testimony plays a significant role in those
disputes.37
33. See Wouter G. Werner, The Politics of Expertise: Applying Paradoxes of Scientific
Expertise to International Law, in THE ROLE OF “EXPERTS” IN INTERNATIONAL AND EURO-
PEAN DECISION-MAKING PROCESSES: ADVISORS, DECISION MAKERS OR IRRELEVANT AC-
TORS? 44, 55 (M. Ambrus et al. eds., 2014) (“Legal experts themselves . . . are also
increasingly dependent on experts form other disciplinary fields. The fragmentation of inter-
national law has created functional subfields where legal norms can often only be applied on
the basis of scientific expert knowledge.”); FOSTER, supra note 3, at 114 (“A significant devel-
opment in the taking of scientific evidence by international courts and tribunals is the system
that has developed in the WTO. The WTO system for taking expert evidence was devised as
a response to the needs of the dispute resolution process in cases involving complex scientific
questions.”).
34. See, e.g., Mark F. Rosenberg & Michael A. Cheah, Arbitrating Environmental Dis-
putes, 16 ICSID REV. 39, 42 (2001) (“[T]here is a body of treaties that deal specifically with
environmental concerns and may provide for arbitration in the event of disagreement be-
tween the signatories.”). Several disputes arising under the 1982 United Nations Convention
on the Law of the Sea Treaty, for example, have relied heavily on expert evidence. See, e.g.,
Phil. v. China, Award, PCA Case Repository 2013-19 (2016) (discussing the terms and neces-
sary skillset of a potential appointed expert); Bay of Bengal Maritime Boundary Arb. (Bangl.
v. India), Award, PCA Case Repository 2010-16 (2014) (appointment of an expert hydrogra-
pher); Guy. v. Surin., Award, PCA Case Repository 2004-04 (2007) (detailing the technical
responsibilities of the appointed expert).
35. See, e.g., OFF. OF THE U. S. TRADE REPRESENTATIVE: THE TRANS-PAC. PARTNER-
SHIP art. 28.1.5, https://ustr.gov/sites/default/files/TPP-Final-Text-Dispute-Settlement.pdf (last
visited Feb. 5, 2017) (“Role of Experts: At the request of a disputing Party, or on its own
initiative, a panel may seek information and technical advice from any person or body that it
deems appropriate, provided that the disputing Parties agree and subject to any terms and
conditions agreed by the disputing Parties. The disputing Parties shall have an opportunity to
comment on any information or advice obtained under this Article.”); Agreement Between
Canada and the Republic of Peru for the Promotion and Protection of Investments art. 42;
Can.-Peru, entry into force June 20, 2007, 2007 Can. T.S. No 10 (2006); Treaty Between the
United States and the Oriental Republic of Uruguay Concerning the Encouragement and
Reciprocal Protection of Investment art. 32, U.S.-Uru., (Nov. 4, 2005), S. Treaty Doc. No.
109-9; United States-Singapore Free Trade Agreement art. 15.23, U.S.-Sing., May 6, 2003,
P.L. 108-78.
36. See, e.g., Agreement on the Application of Sanitary and Phytosanitary Measures,
33 I.L.M. 15 (1994), art. 11.2 [hereinafter SPS Agreement] (“In a dispute under this Agree-
ment involving scientific or technical issues, a panel should seek advice from experts chosen
by the panel.”).
37. See FOSTER, supra note 3, at xiii, xiv, xvii.
Together, these developments have significantly increased the role

that experts play in resolving international disputes, in a broad range of
disputes and a wide variety of forums. Where disputes were once focused
on primarily legal issues with relatively simple fact patterns—the propriety
of war-time actions and expropriations of relatively modest industrial fa-
cilities38—disputes today often involve complex analyses of health, safety,
or environmental risks, or of the economic fates of enormous businesses.
Economists and accountants opine on the amount of damages owed,39
scientists play prominent roles in evaluating the basis of environmental
and health regulations,40 and engineers opine on the operation of indus-
trial facilities and extractive concessions.41 In international criminal cases,
historians and anthropologists testify on the social, political, and historical
contexts of armed conflict and violence.42 And many disputes require the
piecing together of testimony from multiple experts. An investment dis-
pute involving an industrial project might require an engineer to evaluate
the project’s output capacity, an economist to evaluate income streams,
and an accountant and tax specialist to assess the associated regulatory
and business costs.43
The complexity of international disputes means that when decision
makers rely on expert evidence, they rely heavily on it.44 Expert evidence
38. See, e.g., notes 15-20 and accompanying text.

39. See, e.g., Starrett Housing Corp. v. Iran, Interlocutory Award No. ITL 32-24-1, 4
Iran-U.S. Cl. Trib. Rep. 122, 157 (1983).
40. See, e.g., notes 61-66 and accompanying text.
41. See, e.g., Harza v. Iran, Interlocutory Award No. ITL 14-97-2, 2 Iran-U.S. Cl. Trib.
Rep. 68, 71 (1983) (appointing experts in geology, hydraulics, and business practices and
administration of consulting engineering contracts).
42. See Doris Buss, Expert Witnesses and International War Crimes Trials: Making
Sense of Large-Scale Violence in Rwanda, in NARRATIVES OF JUSTICE IN AND OUT OF THE
COURTROOM 26 (Dubravka Zarkov & Marlies Glasius eds., 2014); see also Sean D. Murphy,
The Experience of the Eritrea-Ethiopia Claims Commission 106 PROCEEDINGS OF THE AN-
NUAL MEETING (AM. SOC’Y INT’L L.) 237, 240 (2012) (noting that the Eritrea-Ethiopia
Claims Commission relied on expert witnesses including “psychiatrists, medical doctors, re-
tired military officers from third states, and various military and explosives experts” (cita-
tions omitted)); Avi Singh, Expert Evidence, in PRINCIPLES OF EVIDENCE IN INTERNATIONAL
CRIMINAL JUSTICE 600 (K. Khan et al. eds., 2010) (“As in domestic criminal proceedings,
expert witnesses are critical, seemingly more than eye witnesses, to adduce evidence in inher-
ently complex international trials.”).
43. See Sergey Ripinsky & Kevin Williams, Damages in International Investment Law
175 (2008) (noting that it may be “sometimes essential” to appoint experts with backgrounds
in accounting, finance, and “the industry or market in which the company operates” in valu-
ing businesses).
44. See FOSTER, supra note 3, at 77 (“The complexity of the science requires heavy
consultation with experts, and considerable reliance on their testimony.”); Thomas Cottier,
Risk Management Experience in WTO Dispute Settlement, in GLOBALIZATION AND THE EN-
VIRONMENT: RISK ASSESSMENT AND THE WTO 41, 51 (D. Robertson & A. Kellos eds., 2001)
(arguing that, in EC—Hormones, Australia—Salmon, and Japan Agricultural Products II,
“panels strongly relied upon expert views, and could not have rendered their assessment and
recommendation without such independent advice of individual experts”); RIPINSKY & WIL-
LIAMS, supra note 43, at 174 (characterizing the quantification of damages as a “very complex
often has wide-ranging and significant influence on decision makers and

may even determine the outcome of key issues in a dispute.45 That is true
when the governing legal instruments refer to inherently technical con-
cepts, such as the World Trade Organization’s (WTO) Agreement on the
Application of Sanitary and Phytosanitary Measures (SPS Agreement)
and its references to risk assessments, but also when the legal instruments
do not.46 Expert evidence in international disputes may inform the mean-
ing of terms in international agreements—where the legal text refers to
the reasonableness or necessity of certain actions, for example—and shape
the way that courts and tribunals view the policy choices of states in every
area of governance.47 Even an expert’s “back of the envelope” conclusions
exercise requiring special knowledge,” making “the involvement of valuation experts in arbi-
tral proceedings practically inevitable” and giving those experts a “singularly important
role . . . in the matter of assessing compensation”); Dana Freyer, Assessing Expert Evidence,
in Leading Arbitrators’ Guide to International Arbitration 429, 440 (2008) (“[T]he expert
with technical expertise . . . may effectively act as the final authority on matters of great
importance to the case.”); Miles, supra note 31, at 155-56 (“It is the sum of these incidences
[that have led to the increasing complexity of international disputes] . . . that has wider impli-
cations . . . for . . . the proposition that science and technology are influencing the develop-
ment of international law in fundamental ways, that they play an increasingly pivotal role in
trade and investment regulation and dispute settlement an d, more specifically, that they will
also be central to the treatment of climate-related issues within international economic
law.”); Dave, supra note 28, at 817 (“[A]rbitrators and even lawyers have had to become
proficient in weighing competing expert evidence.”).
45. See Miles, supra note 31, at 160 (“In the context of international litigation, it is the
use of scientific expert evidence to support the legal arguments presented that can be particu-
larly influential on decision-makers and, indeed, determine the outcome of a dispute.”); id.
(“Inevitably, . . . in health and or environment-related cases, there is a heavy reliance on the
input of experts as witnesses or as advocates themselves.”); id. at 161 (“[S]cience-based evi-
dence, its presentation[,] and the authority of the individual experts will be central to a
court’s decision.”); Catherine E. Foster, New Clothes for the Emperor? Consultation of Ex-
perts by the International Court of Justice, 5 J. INT’L DISPUTE SETTLEMENT 139, 140 (2014)
(“[E]xperts wield potentially significant influence in the determination of the scope of sover-
eign States’ future freedoms.”); MICHELLE T. GRANDO, EVIDENCE, PROOF, AND FACT-FIND-
ING IN WTO DISPUTE SETTLEMENT 340 (2009) (“[WTO] panels . . . usually accord
[expert] . . . opinions great weight.”); id. (arguing that the panel in EC — Hormones “relied
entirely on the opinion of the experts to support its finding” that naturally occurring hor-
mones are comparable to hormones administered for therapeutic or zootechnical purposes
and therefore subject to the disciplines of Article 5.5 of the SPS Agreement); id. at 341 (con-
cluding that the Japan-Apples panel “relied entirely upon the opinion of the experts in ad-
dressing” whether a class of apples could act as a disease-transmission vector); see also
United States—Continued Suspension of Obligations in the EC-Hormones Dispute, Appel-
late Body Report, WT/DS320/AB/R (2008), ¶ 480 (“Experts appointed by a panel can signifi-
cantly influence the decision-making process.”).
46. See GRANDO, supra note 45, at 340 (“In most [WTO] cases where experts were
consulted, particularly those brought under the SPS Agreement, the experts’ opinions had a
clear impact on the findings of the panels.”).
47. See Miles, supra note 31, at 77-78
The insights offered by a scientific expert will help determine the application of a
legal concept such as ‘necessity’ or ‘reasonableness’ in the case at hand – and in the
course of this process the scientific expert’s advice will also come to shape the
development of the conventions and usages on which the established meaning of
made on the witness stand may carry significant weight with a court or
tribunal.48 Expert reports, exhibits, and testimony contribute significantly
to the tens of thousands of pages that routinely pile up in international
disputes.49 This trend shows no signs of stopping.50
The use of experts in international law has been especially prevalent
in trade and investment disputes. In the more than fifty years of the multi-
lateral trading system under the 1947 General Agreement on Trade and
Tariffs (GATT), only one panel requested an expert opinion51 out of ap-
proximately 300 disputes,52 and in that instance did not rely on it.53 Early
cases in the WTO also tended to avoid reliance on expert evidence.54 Yet,
today reliance on expert evidence has become a standard practice in the
the legal concept of ‘necessity’ or ‘reasonableness’ in the context of the provision

in question will be based.
48. See FOSTER, supra note 3, at 120; see also Joost Pauwelyn, Expert Advice in WTO
Dispute Settlement, in TRADE AND HUMAN HEALTH AND SAFETY 235, 248 (George A. Ber-
mann & Petros C. Mavroidis eds., 2006).
49. See Lucy Reed, The Kaplan Lecture 2012: Arbitral Decision-Making: Art, Science
or Sport? 12 (2012), http://.arbitration-icca.org/media/1/13581569903770/reed_tribunal_deci
sion-making.pdf (“With flexible rules and no page limits in international arbitration . . . we
see records with 500-page memorials, scores of witness statements and expert reports, and
thousands of exhibits. How can we expect the arbitrators to find a trail up such a mountain
of a record?”); see also CHESTER BROWN, A COMMON LAW OF INTERNATIONAL ADJUDICA-
TION 112 (2007) (noting that parties to ICJ disputes submit “large volumes of complex techni-
cal, scientific or financial material”); FOSTER, supra note 3, at 108 (noting “the mass of
scientific evidence” in the ICJ’s Gabèı́kovo-Nagymaros case); Christophe Bonneuil & Les
Levidow, How Does the World Trade Organization Know? The Mobilization and Staging of
Scientific Expertise in the GMO Trade Dispute, 42 SOC. STUD. SCI. 75, 77 (2012) (citing the
more than 1,000 pages of “discussions” regarding expert evidence in Panel Report, EC —
Approval and Marketing of Biotech Products, WTO/DS291 (2006)).
50. See FOSTER, supra note 3, at 132 (“Expert evidence is likely to be given a higher
profile in international judgments and awards in future.”); Foster, supra note 45, at 139 (“Dis-
agreements over science feature with growing regularity in international disputes.”).
51. See Panel Report, Thailand — Cigarettes, ¶ 27, WTO Doc. DS10/R - 37S/200 (Oct.
5, 1990) (requesting advice from the World Health Organization); see also Pauwelyn, supra
note 48 at 235 (George A. Bermann & Petros C. Mavroidis eds., 2006) (noting that there has
only been “one single resort to experts by GATT 1947 panels”).
52. WTO in Brief: Part 3 – The WTO Agreements, WORLD TRADE ORGANIZATION, 5
(2014), https://.wto.org/english/res_e/doload_e/inbr_e.pdf.
53. See Gruszczynski, supra note 32, at 16–17.
When deciding the [Thailand – Cigarettes] case, the GATT panel took for granted that
the Thai measure was backed up by sufficient scientific evidence. At the same time, it did not
address the claims of Thailand that US cigarettes were actually more harmful than Thai ones
(owing to the presence of additives and the specific process used in their production).
54. Id. at 17-18 (citing US-Gasoline and US- Shrimp as examples of the WTO Appel-
late Body and panels avoiding reliance on expert scientific evidence).
WTO,55 which since 1994 has routinely relied on such evidence.56 And
WTO observers have called for more frequent use of experts in particular
areas, including economics.57
After its initial reticence, the WTO has consistently relied on scientific
experts, in part because a number of WTO provisions expressly require
consistency with scientific norms.58 Panels have in particular relied heav-
ily on scientific expertise in claims submitted under the SPS Agreement59
and the Agreement on Technical Barriers to Trade, and when analyzing a
number of exceptions to WTO obligations, such as those in the General
Agreement on Tariffs and Trade, and the General Agreement on Trade in
Services.60 No international legal text goes further in expressly relying on
technical evidence than the SPS Agreement, which relies in significant
part on science to distinguish between product health measures that are
consistent with the Agreement and those that are not.61 The Agreement
55. See Jessica Lawrence, The Structural Logic of Expert Participation in WTO Deci-
sion-Making Processes, in THE ROLE OF “EXPERTS” IN INTERNATIONAL AND EUROPEAN DE-
CISION-MAKING PROCESSES: ADVISORS, DECISION MAKERS OR IRRELEVANT ACTORS? 173,
176 (M. Ambrus et al. eds., 2014) (“Reliance on scientific expertise has become a standard
practice at the WTO.”); FOSTER, supra note 3, at 133 (citing Japan – Measures Affecting the
Importation of Apples and United States – Import Prohibition of Certain Shrimp and Shrimp
Products as two examples of reliance by WTO panels on expert evidence); Pauwelyn, supra
note 48, at 235 (explaining that, in the first six years of the WTO’s existence, “six panels
appointed scientific experts,” “[t]wo panels requested expert advice from other international
organisations,” and “one panel appointed a linguistic expert” and that “very often parties to
a WTO dispute also nominate experts on their delegation, be they lawyers, economists, scien-
tists or linguists” and NGOs, industry, and academics “have pressed their (expert) opinion on
WTO panels”).
56. See Pauwelyn, supra note 30, at 325 (noting 12 cases between 1994 and 2002).
57. See, e.g., GRANDO, supra note 45, at 342 (“The significant role which experts have
played in scientific disputes suggests that more consideration should be given to the idea of
appointing economic experts when panels are asked to make complex economic determina-
tions that fall outside their area of expertise.”); Andre Sapir & Joel Trachtman, Subsidization,
Price Suppression, and Expertise: Causation and Precision in Upland Cotton, 7 WORLD
TRADE REV. 183, 205–07 (2008) (arguing that panels should use economic experts in cases
requiring complex economic determinations).
58. See Lawrence, supra note 55, at 176 (“[C]onformity with scientific norms is an
explicit criterion of compliance with a number of WTO obligations.”); Gruszczynski, supra
note 32, at 23 (“[S]cience is an important element in all health and environment-related trade
disputes, irrespective of which agreement they are brought under.”); see also Lawrence,
supra note 55, at 174 (“[e]xperts are not something that can be ‘added to’ or ‘subtracted
from’ the WTO in order to tweak input or output legitimacy. Instead their [inputs] . . . are
the fabric from which the organization is cut.”); GRANDO, supra note 45, at 338 (2009) (“In
[WTO] cases where technical matters are at issue, experts may plan an important role in
ensuring an accurate decision.”).
59. See Gruszczynski, supra note 32, at 22 (“[S]cience has become a central benchmark
for the assessment of SPS measures.”).
60. See Pauwelyn, supra note 48, at 235.
61. See Vern R. Walker, Keeping the WTO from Becoming the World Trans-Science
Organization: Scientific Uncertainty, Science Policy, and Factfinding in the Growth Hormones
Dispute, 31 CORNELL INT’L L.J. 251, 253 (1998) (“The central strategy of the SPS Agreement
is to use science to distinguish between those sanitary measures consistent with the Agree-
ment and those in violation of the Agreement.”); see also Panel Report, EC — Approval and
helps to address the possibility that a WTO member might use health con-
cerns as a pretense to impose restrictions on imported goods in order to
protect its own industry from foreign competition.62 Article 5, for exam-
ple, requires governments to ensure that certain health and safety require-
ments are “based on an assessment, as appropriate to the circumstances,
of the risks to human, animal or plant life or health, taking into account
risk assessment techniques developed by the relevant international
organizations.”63
The increasing number of investor-state arbitrations under bilateral
investment treaties and free trade agreements in the last twenty years has
contributed significantly to the overall trend in expert evidence. This sig-
nificant reliance is nowhere more apparent than in valuation determina-
tions in investment arbitrations.64 But experts routinely play prominent
roles in evaluating the scientific basis of health and environmental regula-
tion in investment disputes, too. Challenged regulations have included fuel
additive bans,65 plain packaging requirements for cigarettes,66 public an-
nouncements to discourage drinking contaminated waters,67 and restric-
tions on pesticides.68
Marketing of Biotech Products, Report of the Panel, Annex , Replies by the Scientific Experts,
WT/DS291/R/Add.6 (January 2008) at 170 (providing an example of the technical evidence
that the SPS Agreement relies on when defining risk assessment).
62. See Vern R. Walker, Transforming Science into Law: Transparency and Default
Reasoning in International Trade Disputes, in RESCUING SCIENCE FROM POLITICS: REGULA-
TION AND THE DISTORTION OF SCIENTIFIC RESEARCH 165, 167–68 (Wendy Wagner & Rena
Steinzor eds., 2006).
63. Agreement on the Application of Sanitary and Phytosanitary Measures, art. 5.1-
5.2, Apr. 15, 1994, 1867 U.N.T.S. 493.
64. See IRMGARD MARBOE, CALCULATION OF COMPENSATION AND DAMAGES IN IN-
TERNATIONAL INVESTMENT LAW 181 (2009) (“It is increasingly accepted that employing ex-
perts usually helps the tribunal better understand the complexities involved in calculating
compensation and damages, even though this increases the costs of resolving the dispute and
may slow the process.”); José E. Alvarez, Are International Judges Afraid of Science?: A
Comment on Mbengue, 34 LOY. L.A. INT’L & COMP. L. REV. 81, 86 (2011) (“Many, perhaps
most, [investor-state cases under ICSID] involving sophisticated investors in complex, on-
going enterprises require heavy expert-laden assessments of fair market or going concern
value.”); Joshua B. Simmons, Valuation in Investor-State Arbitration: Toward a More Exact
Science, 30 BERKELEY J. INT’L L. 196, 198 (2012) (“The question of fair market value poses
notable challenges for arbitrators because it relates more closely to finance than law. These
challenges loom large because arbitrators frequently must determine fair market value in
investor-state arbitration.”).
65. See, e.g., Methanex Corp. v. United States, Final Award of the Tribunal on Jurisdic-
tion and Merits, ¶ 101 (NAFTA Ch. 11 Arb. Trib. 2005), http://.italaw.com/sites/default/files/
case-documents/ita0529.pdf.
66. See, e.g., Philip Morris Brands Sàrl v. Oriental Republic of Uruguay, ICSID Case
No. ARB/10/7, Award (2016).
67. See, e.g., Azurix Corp. v. Argentine Republic, ICSID Case No. ARB/01/12, Award
(2006).
68. See, e.g., Chemtura Corp. v. Canada, Award (NAFTA Ch. 11 Arb. Trib. 2010),
http://.italaw.com/sites/default/files/case-documents/ita0149_0.pdf.
Even when a tribunal or court is reluctant to rely on highly technical

evidence, reliance can prove unavoidable. One example is the 2014 ICJ
case Whaling in the Antarctic, in which Australia argued that Japan had
breached an international whaling convention by engaging in a whaling
program that permitted lethal sampling of three species of whales.69 Japan
countered that the lethal sampling fell within an exception in the conven-
tion for whaling conducted “for purposes of scientific research.”70 The
ICJ went to great lengths to avoid expressing an opinion on the scientific
merits of the whaling program but nonetheless went on to determine
“whether the design and implementation of [the program were] reasona-
ble in relation to achieving [the program’s] stated research objectives.”71
That led to an analysis of the reasonableness of the program’s lethal sam-
ple sizes,72 the availability of nonlethal sampling methods,73 and the feasi-
bility of achieving the stated scientific objectives given the sample sizes
and methods.74 On those questions, the Court could not help but wade
into the science and the competing testimonies of the parties’ experts.75 In
the end, a majority of the Court concluded that the design and implemen-
tation of the program was unreasonable in relation to its stated objectives
and thus did not fall within the exception for whaling conducted “for pur-
poses of scientific research.”76
Partly because of the discomfort with expert evidence on display in
Whaling in the Antarctic, the ICJ remains a conspicuous outlier in the
modern trend toward increased reliance on experts.77 Since the 1949
69. Whaling in the Antarctic (Austl. v. Japan: N.Z. intervening), Judgment, 2014 I.C.J.
Rep. 226, ¶ 48 (Mar. 31).
70. Id. ¶ 49.
71. Id. ¶¶ 88, 127.
72. Id. ¶¶ 181, 193, 195-196.
73. Id. ¶ 144.
74. Id. ¶ 212.
75. See Daniel Peat, The Use of Court-Appointed Experts by the International Court of
Justice, 84 BRIT. .B. INT’L L. 271, 287-88 (2014) (“[A]lthough the Court [in Whaling in the
Antarctic] aimed to avoid review of the scientific merits . . . by assessing the reasonableness of
sample sizes, sampling method, and other aspects of the programme, it did just that.”); Marco
Roscini, Evidentiary Issues in International Disputes Related to State Responsibility for Cyber
Operations, 50 TEX. INT’L L.J. 233, 263-64 (2015) (“In the Whaling in the Antarctic case . . .
the experts called by both Australia and Japan gave evidence as expert witnesses and were
cross-examined, and the Court relied heavily on their statements to conclude that the special
permits granted by Japan for the killing, taking, and treatment of whales had not been
granted ‘for purposes of scientific research.’ ”).
76. Whaling in the Antarctic (Austl. v. Japan: N.Z. intervening), Judgment, 2014 I.C.J.
Rep. 226, ¶ 227 (Mar. 31).
77. See FOSTER, supra note 3, at 121; Joost Pauwelyn, The Use, Non-Use and Abuse of
Economics in WTO and Investment Litigation, in 43 WTO LITIG., INV. ARB., & COM. ARB.,
GLOBAL TRADE L. SERIES 172 (Jorge A. Huerta-Goldman, Antoine Romanetti, et al. eds.,
2013) (noting the ICJ’s “continued reluctance to engage with expert evidence”); see also The
Gabèı́kovo-Nagymaros Project (Hung. v. Slovk.), Merits, 1997 I.C.J. Rep. 7, ¶ 54 (Merits)
(Sept. 25) (concluding that the Court need not determine which of the parties’ competing
scientific views were more reliable).
Corfu Channel case, the Court has relied on expert evidence in only two
cases: the Gulf of Maine boundary dispute case in 198478 and the Whaling
in the Antarctic case in 2014.79 Dissenting and separate opinions have long
lamented the Court’s reluctance to rely on expert evidence,80 and the issue
of expert evidence—“how to deal with . . . technical and scientific is-
sues”—has been characterized as “one of the biggest problems” facing the
Court today.81
II. PREVAILING PRACTICES FOR EVALUATING EXPERT EVIDENCE

While it is perhaps not possible or worthwhile to evaluate dispute res-
olution systems on whether they produce outcomes that are “good” or
“bad,” it is key to the legitimacy of dispute resolution systems that the
parties and public have confidence in the design and functioning of the
system.82 When it comes to expert evidence, international courts and
tribunals have done too little to engender that confidence. Rather, they
have largely failed to adapt to the reality that experts are often the most
critical element of an international dispute, and the increasing reliance on
expert evidence in international disputes has continued despite a lack of
accepted or consistently applied standards or best practices.83
78. Delimitation of the Maritime Boundary in the Gulf of Maine Area (Can. v. U.S.),,
Merits, 1984 I.C.J. Rep. 246, 253, 333, 337 (Oct. 12).
79. Whaling in the Antarctic (Austl. v. Japan), 2014 I.C.J. Rep. 148 (Mar. 31), ¶ 20–21.
In the Gulf of Maine dispute, the Special Agreement between the parties to submit the dis-
pute to the Court specifically required the Court to appoint a technical expert to delineate a
precise boundary line, based on the legal principles and rules as set out by the Court. Delimi-
tation of the Maritime Boundary in the Gulf of Maine Area (Can. v. U.S.), Appointment of
Expert Order, 1984 I.C.J. Rep. 165, 165–66 (Mar. 30); see Gulf of Maine, Merits, 1984 I.C.J.
Rep. 246, ¶ 8, 18 (Oct. 12).
80. See Oscar Chinn (U.K. v. Belg.), Merits, 1934 P.C.I.J. (ser. A/B) No. 63, 107–09
(Dec. 12) (individual opinion by Anzilotti, J.); Temple of Preah Vihear (Cambodia v. Thai.),
Judgment, 1962 I.C.J. Rep. 6, ¶ 55 (June 15) (dissenting opinion by Koo, J.); Military and
Paramilitary Activities in and Against (Nicar. v. U.S.), Judgment, 1986 I.C.J. Rep. 6, ¶¶
132–134 (June 27) (dissenting opinion by Schwebel, J.); Kasikili/ Sedudu Island (Bots. v.
Namib.), Judgment, 1999 I.C.J. Rep. 1045, 1118–19, 1131 (Dec. 13) (separate opinion by Oda,
J.); Maritime Delimitation and Territorial Questions Between Qatar and Bahrain (Qatar v.
Bahr.), Judgment, 2001 I.C.J. Rep. 40, ¶¶ 39–41 (Mar. 16) (dissenting opinion by Bernárdez,
J.); Pulp Mills on the River Uruguay (Arg. v. Uru.), Judgment, 2010 I.C.J. Rep. 14, 108 (Apr.
20) (joint dissenting opinion by Al-Khasawneh, J. and Simma, J.).
81. Riddell, supra note 28, at 230. Riddell cites evidence that the Court, without the
knowledge of the parties or the public, does “frequently have recourse to expert advice” in
the form of temporary staff members of the Court’s Registry.” Id. at 240.
82. See RICHARD A. POSNER, HOW JUDGES THINK 3 (2008)
Many of the decisions that constitute the output of a court system cannot be shown
to be either ‘good’ or ‘bad,’ whether in terms of consequences or of other criteria,
so it is natural to ask whether there are grounds for confidence in the design of the
institution and in the competence and integrity of the judges who operate it.
83. See, e.g., Singh, supra note 42, at 600 (“The reliance upon and widespread usage of
expert witnesses [in international criminal cases] has progressed in the absence of clear rules
and regulations with regard to the what and how of expert witnesses.”).
Existing practices do not provide a sufficient framework to evaluate

the reliability of expert evidence. Those practices primarily focus on en-
hancing the rigor of the adversarial process, appointing “independent” ex-
perts, and improving ethical transparency through the disclosure of
potential conflicts of interest. Although these practices may sharpen
points of disagreement and highlight flagrant biases, and the evaluation of
expert evidence has generally improved over time,84 none of those prac-
tices assist the decision maker in engagement with the substance of that
evidence. Robust and transparent reliability analysis remains the excep-
tion, not the rule.
Although today’s practices in handling expert evidence reflect some
minor improvement over nineteenth and early twentieth century practices,
they at their core have remained largely unchanged for many years.85 It is
true, for example, that some early procedures limited parties to written
responses to expert testimony.86 Yet the ability of a party to question ex-
perts at a hearing, which today is enshrined in most procedural rules,87 is a
longstanding practice, not an innovation prompted by the heightened im-
portance of this sort of evidence. The 1936 statute of the Permanent
Court of International Justice, for example, provided for the examination
84. Pauwelyn, supra note 77, at 171 (“Ten years later [after the WTO’s report in EC -
Bananas], the sophistication and quality of analysis in both damage and retaliation calcula-
tions have advanced tremendously.”).
85. See FOSTER, supra note 3, at 79 (describing “traditional approaches” in interna-
tional courts and tribunals as including those “that allow the parties to present countervailing
scientific evidence, through expert reports, affidavits, advocacy and the appearance and ex-
amination of the parties’ own expert witnesses”).
86. See e.g. U.S. DEP’T OF STATE, FOREIGN RELATIONS OF THE UNITED STATES 1902:
WHALING AND SEALING CLAIMS AGAINST RUSSIA, APPENDIX I, at 428 (1903) (“The Arbitra-
tor observed that the hearing of the experts by the arbitrator could have taken place without
the presence of the other party.”); see also Costa Rica Packet Arbitration (Gr. Brit. v. Neth.),
184 C.T.S. 240 (1897).
87. See, e.g., IBA RULES ON THE TAKING OF EVIDENCE IN INTERNATIONAL ARBITRA-
TION, art. 6(6) (Int’l Bar Ass’n [IBA], May 29, 2010) (providing that a tribunal-appointed
expert “may be questioned . . . on issues raised in his or her Expert Report, the Parties’
submissions or Witness Statement or the Expert Reports made by the Party-Appointed Ex-
perts.”); id. at art. 8(3)(b) (“With respect to oral testimony at an Evidentiary Hearing . . .
following direct testimony, any other Party may question such witness.”); id. at art. 8(3)(d)
(providing that a tribunal-appointed expert “may be questioned by the Parties or by any
Party-Appointed Expert on issues raised in the Tribunal-Appointed Expert Report, in the
Parties’ submissions or in the Expert Reports made by the Party-Appointed Experts”); IN-
TERNATIONAL DISPUTE RESOLUTION PROCEDURES, art. 25.4 (Int’l Centre for Dispute Reso-
lution [ICDR], June 1, 2014); INT’L CHAMBER OF COMMERCE RULES OF ARBITRATION, art.
25.4 (Int’l Chamber of Commerce [ICC], Jan. 1, 2012); see also ALAN REDFERN & MARTIN
HUNTER, LAW AND PRACTICE OF INTERNATIONAL COMMERCIAL ARBITRATION 312 (4th ed.
2004) (“If each party presents conflicting evidence of technical opinion, the expert witnesses
must be prepared to appear in person before the arbitral tribunal for examination. Other-
wise, the arbitral tribunal will have no means of evaluating the weight that should be given to
the opinions presented by one side or the other.”); ARBITRATION RULES, art. 21.4 (London
Court of Int’l Arbitration [LCIA], 2014).
of expert witnesses by the parties and the court88 and the same provision
remains in force under the 1946 ICJ rules.89 The rules governing other
twentieth century tribunals and commissions generally permitted parties
to examine expert witnesses and comment on their reports.90 It has long
been true, then, that “the communication of the expert’s report . . . to the
parties and the grant of adequate opportunity . . . to comment upon it and
to question the expert” are “fundamental procedural safeguards.”91
As a way to improve upon this longstanding core mechanism for han-
dling expert evidence, commentators have suggested that decision makers
should more frequently appoint their own experts, rather than rely on
party-appointed experts.92 But the authority of a tribunal to appoint its
own expert, alone or in addition to any experts that the parties might of-
fer, is nothing new.93 Similarly, experts have long been required to dis-
close their background, qualifications, and experience, according to a
decision maker’s order or other rule.94
88. Statute for the Permanent Court of International Justice art. 51, Dec. 16, 1920, 6
L.N.T.S. 407 (1921) [hereinafter P.C.I.J. Statute]; see also WHITE, supra note 9, at 39.
89. Statute of the International Court of Justice art. 51, June 26, 1945, 59 Stat. 1062
[hereinafter I.C.J. Statute].
90. See, e.g., Rules of Procedure of the German Arbitral Commission, Rule 45(a), in 1
DECISIONS OF THE ARBITRAL COMMISSION ON PROPERTY, RIGHTS, AND INTEREST IN GER-
MANY 192, 200 (1958); CHARLES N. BROWER & JASON D. BRUESCHKE, THE IRAN-UNITED
STATES CLAIMS TRIBUNAL 201 n.957 (1998) (citing Iran-U.S. Claims Tribunal cases and ob-
serving that “[t]he proceedings may involve several phases: the expert’s investigation; sub-
mission of a draft report; comments by the parties on the draft report; submission of the
expert’s final report; comments by the parties on the final report; and a hearing on the ex-
pert’s final report”); Corfu Channel (U.K. v. Alb.), Merits, 1949 I.C.J. Rep. 4, 7 (Apr. 9)
(noting that each party cross-examined the expert witnesses of the other party).
91. WHITE, supra note 9, at 81 (making this statement in 1965).
92. FOSTER, supra note 3, at 79-80
We should endorse a move towards processes for consultation of experts that draw
on a blend of investigative and adversarial procedures. . . . A helpful starting point
may be to hold an ‘organisational conference’ at the outset of proceedings in order
to clarify expectations and requests regarding the procedures to be followed in
relation to evidence and proof.
93. See, e.g., Iran-U.S. Claims Tribunal, Final Tribunal Rules of Procedure, art. 27, in 2
Iran-U.S. Cl. Trib. Rep. 405, 430 (May 3, 1983); G.A. Res. 31/98, Arbitration Rules of the
United Nations Commission on International Trade Law [UNCITRAL], art. 27 (Dec. 15,
1976); I.C.J. Statute art. 50, June 26, 1945, 59 Stat. 1062; P.C.I.J. Statute art. 50, Dec. 16, 1920,
6 L.N.T.S. 407 (1921); Convention for the Pacific Settlement of International Disputes art.
90, Oct. 18, 1907, 36 Stat. 2199; see also BROWER & BRUESCHKE, supra note 90, at 199-200,
202 (noting that the Iran-U.S. Claims Tribunal, under the UNCITRAL Rules, has the author-
ity to appoint experts, in addition to the experts that the parties may appoint, although the
Tribunal “has appointed experts in only eight of the 960 large claims”).
94. See, e.g., Starrett Housing Corp. v. Iran, Final Award No. 314-24-1, 16 Iran-U.S. Cl.
Trib. Rep. 112, 196–97 (1987) (“In determining the weight to be given to the Expert’s Report,
the Tribunal must first consider his qualifications. . . . The Tribunal . . . reviewed the Expert’s
background and experience before appointing him.”); see also Robert Wisner et al., Effective
Use of Economic Experts in International Arbitration: Counsel’s Role and Perspective, in EU
AND US ANTITRUST ARBITRATION: A HANDBOOK FOR PRACTITIONERS 237, 242-43 (Gordon
Blanke & Phillip Landolt eds., 2011).
Other efforts to improve the handling of expert evidence have focused

on two procedural aspects: enhancing the disclosure of potential conflicts
of interest95 and encouraging cooperation and engagement among party-
appointed experts with the aim of narrowing the scope of disagreement.96
For example, to further witness cooperation the IBA Rules on the Taking
of Evidence in International Commercial Arbitration contemplate that
tribunals might require the parties’ expert witnesses to meet and discuss
their reports before the hearing and to develop a joint report summarizing
the points of agreement and disagreement.97 The IBA Rules also contem-
plate that decision makers might enhance the adversarial process through
“witness conferencing” or “hot-tubbing,” in which they may hear more
than one expert witness at the same time and encourage the experts to
comment on the other experts’ views.98 Under this approach, party-ap-
pointed witnesses present their views and respond to the tribunal’s ques-
tions.99 Another proposed variation of the approach is the “Sachs
Protocol,” which “envisages at least one ‘open’ session at which the parties
counsel may address and [ ] question the expert team in the presence of
the arbitral tribunal.”100
Existing efforts to improve evidentiary reliability each have their own
limitations. Although relying on the adversarial process may assist in ex-
ploring and testing the strength of the expert’s methods and conclu-
sions,101 it may also involve efforts to muddy rather than clarify the issues
95. See Mark Kantor, A Code of Conduct for Party-Appointed Experts in International
Arbitration – Can One Be Found?, in 26 ARB. INT’L 323, 323 (Park ed., 2010); see also
Chartered Inst. of Arb. [CIArb], Protocol for the Use of Party Appointed Expert Witnesses in
International Arbitration art. 4(4)(b) (2007) [hereinafter CIArb Protocol] (requiring experts
to disclose “any past or present relationship with any of the Parties, the Arbitral Tribunal,
counsel or other representatives of the Parties, other witnesses and any other person or entity
involved in the Arbitration”).
96. MARBOE, supra note 64, at 183 (suggesting that investment tribunals should com-
pare expert reports on damages calculations and “identify those aspects of the valuation
where there is agreement and where there is disagreement” or to “request the experts to
develop and submit such lists of agreements and disagreements”).
97. IBA RULES art. 8(3)(f). The CIArb Protocol also provides for discussions among
party-appointed experts to “identif[y] and list[ ] the issues upon which they are to provide an
opinion,” as well as “any tests or analyses which need to be conducted”; “exchange draft
outline opinions for the purposes of these meetings”; and prepare a joint statement identify-
ing areas of agreement and areas of disagreement. CIArb Protocol art. 6(1)(a), (b).
98. IBA RULES art. 8(3)(f); see also Freyer, supra note 44, at 439 (“This practice has
been used extensively in Asia and Australia and is attracting increasing support among Euro-
pean and American arbitrators.”).
99. See id.; FOSTER, supra note 3, at 123.
100. See Klaus Michael Sachs, Experts: Neutrals or Advocates, INT’L COUNCIL FOR
COM. ARB. (May 25, 2010), http://arbitration-icca.org/conferences-and-congresses/
ICCA_RIO_2010/ICCA_RIO_2010_Klaus_Michael_Sachs.html (proposing a variation of ex-
pert-witness conferencing); see also Alison Ross, A Sachs- New Approach to Expert Evi-
dence?, GLOBAL ARB. REV. (May 27, 2010), http://globalarbitrationreview.com/news/article/
28439/a-sachs-new-approach-expert-evidence.
101. See FOSTER, supra note 3, at 79 (“Much can be learnt [through the adversarial
process], as avenues of proof are explored and tested by counsel working with their experts.
at hand, or to cast doubt on an expert’s testimony through cursory analysis

that may have visceral appeal to laymen but lack any technical basis.102
An expert appointed by a decision maker may do nothing more than act as
a de facto member of the court or tribunal whose opinion has decisive
effect,103 or reflect the strategic demands of the parties on their preferred
expert choices.104 And excessive reliance on credentials can short-circuit a
full-fledged examination of expert evidence.105
More fundamentally, existing efforts may serve to improve the proce-
dures used to appoint experts and to evaluate the evidence that they pre-
sent, but they do not directly address the substance of expert evidence.
Part III outlines the consequences of failures to engage comprehensively
and substantively with expert evidence. In the end, no matter the methods
of appointment or the taking of expert evidence, a court or tribunal must
still decide for itself whether the testimony is reliable and how much
weight to afford it.
III. INADEQUACY OF PREVAILING PRACTICES
Although “[i]t is all too easy to agree that international judges do not
handle facts as well as they could,”106 and many international disputes are
resolved in ways that transparently and comprehensively address the relia-
bility of expert evidence, experience continues to demonstrate that, with-
out a comprehensive analytical framework, decision makers often struggle
with the complexities of expert evidence. This Part explores the three pri-
mary difficulties facing decision makers in international disputes when
confronted with expert evidence.
The process of cross-examination is most valuable in testing the strength of propositions re-
lied upon by the parties.”).
102. See FOSTER, supra note 3, at 100-01.
103. See MARBOE, supra note 64, at 182
[T]he costs of such a tribunal-appointed expert may be considerable. Further-

more, the parties may be concerned that the independent expert will become de
facto the fourth, and decisive, arbitrator on the compensation or damages issue. It
appears, therefore, that the continental practice of tribunal-appointed expert is
only in exceptional cases advisable in international investment disputes.
(internal citations omitted).

104. See Bonneuil & Levidow, supra note 49, at 84-86 (observing that, in the WTO EC
— Approval and Marketing of Biotech Products case, the parties employed different strate-
gies to select experts perceived as favorable to their position, one party by flooding the selec-
tion pool with candidates and the other party by proposing only a few candidates and
rejecting all of its adversary’s candidates, and that the Panel “simply accepted the parties’
demands for rejection without evaluating them”).
105. See Wisner, supra note 94, at 246.
106. See Alvarez, supra note 64, at 96.
A. Poorly Reasoned Reliance on Expert Evidence

Experts can play a valuable role in international disputes by gathering
and assessing complex technical facts that are outside the general knowl-
edge of the lawyers who generally sit on tribunals and courts.107 These
experts are intended to “assist, educate[,] and advise the arbitral tribu-
nal.”108 But they may not become the de facto decision makers them-
selves, because international jurists and arbitrators are generally
prohibited from delegating their adjudicatory authority to others.109 And
nothing about specialized testimony makes it inherently reliable or an ob-
jective basis for decision making. Instead, the development—and presen-
tation—of expert knowledge is a process susceptible to societal influences,
errors, and biases.110 For those reasons, careful reliance on— rather than
delegation or outsourcing to—experts is essential to the accuracy and le-
gitimacy of international dispute settlement.
Yet, all too often international courts and tribunals base their deci-
sions on expert evidence without transparently and systematically deter-
mining whether that evidence is a reliable basis for decision. This leaves
them open to criticisms that they have delegated their decision-making
duties to subject matter experts111 and elevated expert testimony from the
107. See RIPINSKY & WILLIAMS, supra note 43, at 174-75 (“The experts’ function is to
assist in ascertaining and assessing facts which involve technical matters outside the general
knowledge of judges or lawyers.”); Nicolas Ulmer, Assessing Damages—Are Arbitrators
Good at It? Should They Be Assisted by Experts? Should They Be Entitled to Decide ex
aequo et bono? Some War Stories, 6 J. WORLD INV. & TRADE 11, 11 (2005) [hereinafter
Assessing Damages] (reasoning that arbitrators “are not generally very good at assessing
damages” because they are “mainly lawyers and law professors,” not “accountants,” “invest-
ment bankers,” or “quantity surveyors”).
108. See Martin Hunter, Expert Conferencing and New Methods, 4 TRANSNAT’L DISP.
MGMT. 1, 1 (2007).
109. See WHITE, supra note 9, at 13 (“The proper administration of international justice
requires that there should be no delegation of the judicial function to anyone outside the
tribunal, unless this is clearly the intention of the parties, to cover some special situation.”);
MARBOE, supra note 64, at 4 (cautioning that, in investment disputes, “the decision about the
amount of compensation or damages must not be delegated to the experts”); RIPINSKY &
WILLIAMS, supra note 43, at 179 (“It is frequently emphasized that tribunals should treat
expert opinions, including those produced by the tribunal-appointed experts, as items of evi-
dence and avoid delegating their decision-making powers to experts.”); Caroline E. Foster,
The Consultation of Independent Experts by International Courts and Tribunals in Health and
Environment Cases, 20 FINNISH Y.B. INT’L L. 391, 395 (2009) (recognizing the “need for
international courts and tribunals to be alert to the possibility of inadvertent delegation to
experts”).
110. See Gruszczynski, supra note 32, at 25
The construction of scientific knowledge is a social process that is culturally depen-

dent, and the assessment of risks relies on a number of normative non-scientific
approximations and assumptions (e.g.[,] owing to gaps in the knowledge, with
some of them being irreducible). This means that science is not capable of being a
fully objective benchmark in international trade disputes.
111. See WHITE, supra note 9, at 13 (noting that the “almost insuperable difficulties of
proving that a tribunal had not made an independent evaluation of some, or even all, of the
relevant facts, but had followed the evaluation of its expert, can be readily imagined” and
status of evidence to “irrefutable truth.”112 Failure to assess the reliability

of expert evidence may result in undue reliance on it, compromising a de-
cision’s legitimacy and inappropriately affecting the regulatory choices of
governments, the legitimate commercial interests of companies, or the fate
of individuals.113
The perception of undue reliance on expert evidence has dogged in-
ternational tribunals and courts. For example, in Corfu Channel, the only
ICJ case in the pre-modern era to have relied on expert evidence, a dis-
senting opinion lamented that the Court’s award of damages “makes
hardly any reference to the many . . . documents accepted as evidence as
damage,” because “something should have been said on their value as evi-
dence,” and the expert’s testimony was accepted without meaningful ex-
amination.114 The post-war Franco-Italian Commission, which relied
frequently on expert evidence, also seemed to do so at times without any
meaningful evaluation of the basis for and quality of expert evidence. For
example, a 1963 decision of the Commission stated that, although “[i]t is
citing the 1949 ICJ decision in Corfu Channel as one that “rel[ied] upon the experts’ analysis
of the facts to a pronounced degree”); Geoffrey Beresford Hartwell, Assessing Damages, 6 J.
WORLD INV. & TRADE 7, 8-9 (2005) (“[I]n legal theory—delegatus non posse delegare—the
tribunal may not delegate its decision-making to the expert. In practice, however, that is
difficult to avoid. . . . [the expert] may . . . have decision-making power which is not always
apparent. It may be that the tribunal will depend on him very substantially.”); Serge Lazareff,
Assessing Damages, 6 J. WORLD INV. & TRADE 17, 18 (2005) (“If we turn to the civil-law
approach, I must say that I am strongly opposed to the systematic appointment of an expert
by the Tribunal because I think it results in delegated justice. It is very difficult for a tribunal
not to follow the expert it has itself appointed.”); Ulmer, supra note 107, at 12 (noting that
arbitrators are very unlikely “to go head-to-head” with certain experts but instead would
“very grateful” to leave all the technical issues to them); Alvarez, supra note 64, at 97 (“We
ought to be equally concerned with whether judges are, in some cases, overly deferential and
only too ready to accept some forms of expert-driven ‘fact.’ ”); Quelques Remarques sur la
Preuve devant la Cour Permanent et la Court Internationale de Justice, 7 ANNUAIRE SUISSE
DE DROIT INT’L 101 (1950), translated in SANDIFER, EVIDENCE BEFORE INTERNATIONAL
TRIBUNALS (1975), at 327 n.144 (“Although the Court, like a national court, is never bound
by the conclusions of the experts, these conclusions, as in municipal law, usually exert a
relatively large influence upon the decisions taken.”); see also Appellate Body Report, US —
Continued Suspension, ¶ 436, WTO Doc. WT/DS320/AB/R (adopted Oct. 16, 2008) (“[T]he
manner in which [experts’] opinions are solicited and evaluated can have a significant bearing
on a panel’s consideration of the evidence and its review of a domestic measure, especially in
cases . . . involving highly complex scientific issues.”).
112. Miles, supra note 31, at 160.
113. See id. (asserting that the complexity of modern international litigation “coupled
with a reluctance to engage in in-depth questioning of . . . scientific evidence” may mean that
“courts can be overly deferential to scientific expertise or that the weight given to individual
written reports could be misplaced”); cf. Alvarez, supra note 64, at 97 (criticizing decisions
that “rely on a form of ‘scientific’ expertise without exploring whether this is truly war-
ranted . . . silently constrain the regulatory autonomy of governments).
114. Corfu Channel (U.K. v. Alb.), Judgment, 1949 I.C.J. Rep. 252, 253 (Apr. 9) (dis-
senting opinion by Eèer, J.). The Court compared the damages estimates of tribunal-ap-
pointed experts to the claimed damages of the United Kingdom, determining that the latter
were reasonable in light of the expert evidence. See Corfu Channel (U.K. v. Alb.), Judgment,
1949 I.C.J. Rep. 244, 249 (Dec. 15).
certain that the opinion of the expert does not bind the Commission,”
there was no reason not to adopt the expert’s conclusion, “unless his argu-
mentation is in contradiction with the facts of record, with the legal provi-
sions or the rules of logic.”115 The written decision does not indicate what,
if any, substantive analysis the Commission undertook in evaluating the
expert’s opinion.116
Poorly reasoned evaluations of expert evidence can result not only in
the wholesale adoption of an expert’s opinion but also in subtler errors.
When decision makers must make numerical estimates, for example, the
initial values put forward by the parties and their experts may “anchor” a
decision maker’s final calculation—establishing a false equivalence be-
tween the competing values of the parties—even if the initial values are
extreme and wholly unreasonable.117
When it comes to calculating damages, the anchoring effect has led to
a common, although perhaps overly broad, perception that international
courts and tribunals do not make fully reasoned judgments when awarding
damages but instead “split the baby” and award damages at the midpoint
of the two sides’ valuations.118 An oft-cited example is Santa Elena v.
Costa Rica, where the tribunal expressly chose to value an expropriated
asset as exactly half of the sum of the two competing amounts submitted
by the parties and their experts, without articulating any analysis of the
evidence supporting those amounts.119 The phenomenon may stem from
a reluctance to engage with the evidence in a meaningful way, because of
115. Heretier de Sar Mgr. le Duc de Guise, Dec. No. 162 (Nov. 20, 1953), 13
U.N.R.I.A.A. 162, 168 (1963), translated in SANDIFER, EVIDENCE BEFORE INTERNATIONAL
TRIBUNALS (1975), supra note 10, at 327.
116. See id.
117. See Chris Guthrie et al., Inside the Judicial Mind, 86 CORNELL L. REV. 777, 793-94
(2001) (“[T]he influence on judges of misleading anchors, such as litigants’ requests for dam-
age awards, can produce biased damage awards.”).
118. See MARBOE, supra note 64, at 182 (“In some cases, the tribunals set the amount of
compensation exactly halfway between the claimant’s and the respondent’s valuations, thus
appearing to ‘split the baby.’ ” (internal citations omitted)); RIPINSKY & WILLIAMS, supra
note 43, at 191 (“[O]n a number of occasions, tribunals appear to have simply ‘split the baby,’
taking the mean between alternative valuations produced by the parties, or were otherwise
opaque about how they arrived at the amount of compensation.”); Joost Pauwelyn, supra
note 77, at 171-72 (“[I]n the not too distant past, WTO arbitrators as well as investor-State
tribunals, rather than using sound economic models and data to calculate damages or permit-
ted trade retaliation, were more or less ‘splitting the difference,’ that is adding up what the
two opposing parties claimed and then awarding half of that amount.”).
119. Compañı́a del Desarrollo de Santa Elena, S.A. v. Republic of Costa Rica, ICSID
Case No. ARB/96/1, Award, ¶¶ 93-95 (Feb. 17, 2000). A number of other decisions, many of
them before the last ten years, appear to have made awards essentially at the midpoint of the
parties’ estimates, see, e.g., Middle East Cement Shipping and Handling Co. S.A. v. Arab
Republic of Egypt, ICSID Case No. ARB/99/6, Award, ¶¶ 150-151 (Apr. 12, 2002) (valuing a
ship); Saghi v. Islamic Republic of Iran, Award, 29 Iran-U.S. Cl. Trib. Rep. 20, 55, ¶ 104
(1993) (noting expressly that the award fell between claimants’ and respondent’s estimates).
its complexity and legal decision makers’ lack of expertise in financial

analytics.120
Although most decision makers may not admit to adopting the mid-
point between competing valuations,121 the pull of the less detectable
anchoring effect may still influence decisions and lead disputing parties—
and their experts—to put forward more extreme valuations as a means to
shape a decision maker’s damages calculation.122 Even decisions that
serve as examples of otherwise robust efforts to analyze the reliability of
expert evidence have included attempts to reconcile the parties’ compet-
ing numbers, indicating that the parties may influence the outcome of the
dispute simply by raising or lowering their initial values. In CME v. Czech
Republic, for example, the tribunal stated expressly that it would attempt
to “close[ ]” the “gap between the experts’ valuations of roughly USD 210
million” through “a rough assessment.”123
B. Poorly Reasoned Dismissal of Expert Evidence

Neglecting to effectively evaluate expert reliability can lead to undue
or wholesale adoption of expert opinion, but it can also have quite the
opposite effect, including the flippant disregard of evidence that is reliable
and that would usefully aid in the decision-making process.124 It is too
often easier for non-specialists to ignore evidence that they consider too
difficult to understand, to cite minor flaws in a technical analysis as a pre-
text for disregarding it, or to create other doubts as to the reliability of
expert evidence, than it is to take the time and effort to understand and
appropriately rely on highly technical material.125 In other instances,
120. RIPINSKY & WILLIAMS, supra note 43, at 122 (noting that reluctance to engage
with the details of valuation may lead to “splitting the baby” or some other means of simpli-
fying the valuation process when decision-makers “get lost in the intricacies of valuation
techniques”); Simmons, supra note 64, at 209 (“Because of their backgrounds, arbitrators
may be reluctant to immerse themselves in the detailed formulas and spreadsheets submitted
by the parties.”).
121. See Christopher R. Drahozal, Behavioral Analysis of Private Judging, 67 L. & CON-
TEMP. PROBS. 105, 118 (2004) (surveying available empirical studies on the anchoring effect—
also known as “extremeness aversion”—and concluding that “[a]rbitrators (at least in com-
mercial cases) do not seem to be subject to extremeness aversion”); Richard . Naimark &
Stephanie E. Keer, Arbitrators Do Not “Split the Baby” – Empirical Evidence from Interna-
tional Business Arbitration, 18 J. INT’L ARB. 573 (2001); see also RIPINSKY & WILLIAMS,
supra note 43, at 191 (noting that, “in more recent arbitral awards, arbitrators have enhanced
the quality of analysis” of expert evidence on valuations).
122. See MARBOE, supra note 64, at 182 (reasoning that the “split the baby” problem
may “encourage the parties to overestimate or underestimate their claims”); Reed, supra
note 49, at 5 (citing the “anchoring” effect in cautioning that “[c]ounsel must think carefully
about the numbers used in estimating damages in initial requests for arbitration”).
123. CME Czech Republic B.V. v. Czech Republic, Final Award, ¶ 596 (UNCITRAL
Arb. Proc. 2003), http://italaw.com/sites/default/files/case-documents/ita0180.pdf.
124. See Miles, supra note 31, at 161 (“[P]erhaps[ ] the risk [in international dispute
resolution for a] is that the [expert] evidence is not given sufficient consideration at all.”).
125. See Lawrence, supra note 55, at 188 (“Because of their lack of training and exper-
tise in particular disciplines, WTO dispute settlement bodies often fail to adequately assess
“egocentric bias” may mean that decision makers rely on their “own
preconceptions and give undue weight to or even seek out evidence to
support those preconceptions, rather than fully consider the parties’
arguments.”126
Decisions throughout the years have illustrated the difficulties that
may result from the poorly reasoned rejection of expert evidence and cur-
sory substitution of a legal decision maker’s lay opinion. Early in its exis-
tence, the Iran-U.S. Claims Tribunal, in Starrett Housing Corp. v. Iran,
appointed a valuation expert, who prepared a detailed report at a cost of
over one million dollars, and permitted the parties to comment in writing
on the report and examine the witness at a hearing.127 Although the Tribu-
nal claimed to rely on the expert’s discounted cash flow (DCF) methodol-
ogy, it essentially ignored the expert’s application of that methodology to
key facts of the dispute.128 Instead, the Tribunal made its own qualitative
and generic assessment, resulting in significant changes to the amount of
damages awarded.129 Judge Holtzmann, in a concurring opinion, argued
that this represented a departure from an otherwise robust assessment of
the expert evidence on valuation.130 For one, the Tribunal rejected the
expert’s use of compound interest,131 despite the “modern economic real-
ity” confirmed by the expert, that investors were required to pay on the
loans that had financed the expropriated project.132
The 1989 decision in the Salmon and Herring v. Canada case, which
arose under the United States-Canada Free Trade Agreement, has been
criticized for the tribunal’s apparent “willingness to substitute [its] own
judgment for the numerical determinations of governmental experts based
on [its] own reading of scientific texts.”133 The United States had chal-
and incorporate scientific contributions.”); Ulmer, supra note 107, at 14 (recounting from an
experience as counsel in a commercial arbitration that “it is, in general, much easier to pick
holes, to create doubts, in the damages presentation to actually put forward an iron-clad
one”).
126. See Reed, supra note 49, at 7.
127. See Starrett Housing Corp. v. Iran, Final Award, 16 Iran-U.S. Cl. Trib. Rep. 112,
120, ¶ 11 (1987).
128. Id. at 117, ¶¶ 337-343 (1987); see generally RIPINSKY & WILLIAMS, supra note 43,
at 176 n.270 (observing that the extensive expert evidence in Starrett Housing went “largely
ignored” by the tribunal).
129. Starrett Housing, 16 Iran-U.S. Cl. Trib. Rep. at 117, ¶¶ 337-343.
130. Starrett Housing, 16 Iran-U.S. Cl. Trib. Rep. at 241–54 (1987) (Holtzmann, J., con-
curring opinion).
131. See Starrett Housing, 16 Iran-U.S. Cl. Trib. Rep. ¶ 370.
132. Starrett Housing, 16 Iran-U.S. Cl. Trib. Rep. at 249-54 (Holtzmann, J., concurring
opinion).
133. See, e.g., David A. Wirth, The Role of Science in the Uruguay Round and NAFTA
Trade Disciplines, 27 CORNELL INT’L L.J. 817, 845 (1994). Although the “experts” in this
case were not testifying witnesses—they were specialists within the Canadian regulatory
agencies responsible for the challenged regulation—the panel’s consideration of their pur-
ported expertise is instructive of the potential for decision makers to dismiss expert evidence
without adequately assessing the reliability of that evidence.
lenged Canadian regulations that required commercial fishing companies

to off-load on Canadian territory one hundred percent of all caught her-
ring and salmon, including fish immediately destined for export.134 The
panel did not express a conclusion that the Canadian government special-
ists were not experts in their field, nor did the panel rely on the counter-
vailing testimony of other experts. Yet the panel reviewed statistics and
fisheries texts on its own and “on the basis of logical analysis” concluded
that Canada’s regulation did not fall within the exception for measures
“relating to the conservation of exhaustible natural resources” because
“reliable sampling data can be obtained without requiring access to 100%
of the catch.”135
Tribunals and courts have been criticized consistently over the years
for dismissing expert evidence without adequate reasoning when tasked
with calculating damages. Many legal decision makers have been reticent
to use “income-based” valuation methods, such as the DCF method, that
are the most widely used in actual business practice,136 although there is
some indication that the method is increasingly accepted in international
disputes, at least in concept.137
Even when a tribunal accepts the methodology as a general matter,
difficulty remains in applying the methodology to the specifics of individ-
ual disputes. In the 2014 decision in Venezuela Holdings v. Venezuela, for
example, the tribunal’s valuation analysis focused on whether the “dis-
count rate,” which translates future cash flows into a present value, should
account for the risk that the host country will confiscate the investment.138
134. In the Matter of Canada’s Landing Requirement for Pacific Coast Salmon and Her-
ring, U.S.-Canada Free Trade Agreement Binational Panel Review, Panel No. CDA-89-1807-
01 (Oct. 16, 1989).
135. Id. ¶¶ 7.08, 7.21, 7.29.
136. See MARBOE, supra note 64, at 206–07.
[I]n the practice of international courts and tribunals, the income approach has
long been regarded with skepticism. The necessary forecasts as well as the dis-
counting process appeared to be hardly understandable for lawyers who are not
specially trained in economic disciplines. . . . This reluctance to apply an income-
oriented valuation approach in international proceedings has been predominant
for decades despite the fact that the income approach is the most widely used
approach in economic reality and, thus, routinely applied by valuation experts in
many different contexts.
MARK KANTOR, VALUATION FOR ARBITRATION 53 (2008) (“The contrast between the pref-
erence among international investment tribunals for Asset-Based methods and the prefer-
ence among valuation organizations and professions for Income-Based methods is striking.”).
137. See MARBOE, supra note 64, at 215.
138. Venezuela Holdings B.V. v. Bolivarian Republic of Venezuela, ICSID Case No.
ARB/07/27, Award, ¶¶ 360–368 (2015). See generally RICHARD A. BREALEY ET AL., PRINCI-
PLES OF CORPORATE FINANCE 16 (8th ed. 2006) (“To calculate present value, we discount
expected payoffs by the rate of return offered by equivalent investment alternatives in the
capital market. This rate of return is the discount rate, hurdle rate, or opportunity cost of
capital.”). Future cash flows are worth less than cash flows today. See id. (“The first basic
principle of finance is that a dollar today is worth more than a dollar tomorrow, because the
After determining that any valuation must account for the risk that Vene-
zuela would confiscate the investment, the tribunal concluded that it was
“unable to adopt the approach used by the Claimants’ expert, which does
not take this risk into account.”139 Venezuela’s experts had taken the con-
fiscation risk into account, calculating a range of discount rates more than
twice the rate offered by the claimants’ expert and greatly reducing the
overall valuation.140 Adopting a discount rate within that range, the tribu-
nal noted only that other tribunals “in circumstances comparable to the
present case” had adopted similar discount rates, including a commercial
arbitration tribunal evaluating contract claims related to the same pro-
ject.141 The tribunal did not evaluate the underpinnings of Venezuela’s
expert evidence whatsoever, nor did it explain why it considered the deci-
sions of other tribunals—all but one considering claims against countries
other than Venezuela—were “in circumstances comparable” to this partic-
ular time period, industry, and economy.142
A larger debate—about whether to ever rely on expert evidence—has
been playing out prominently in the ICJ in recent years. Those opposed to
reliance on expert evidence have been in the majority in most recent cases
but dissenting voices have grown louder and more numerous over the
years.143 A trio of cases illustrates the criticisms the Court has received
dollar today can be invested to start earning interest immediately. Financial managers refer
to this as the time value of money.”).
139. Venezuela Holdings B.V., ICSID Case No. ARB/07/27, Award, ¶ 365.
140. See id. ¶ 366.
141. See id. ¶¶ 367–368.
142. See id.
143. See, e.g., Oscar Chinn (U.K. v. Belgium), Judgment, PCIJ Rep. Series A/B No. 63,
107-09 (individual opinion of Judge Anzilotti); Temple of Preah Vihear (Cambodia v. Thai-
land) (Merits), 1962 I.C.J. Rep. 6, ¶ 55 (dissenting opinion of Judge Wellington Koo); Mili-
tary and Paramilitary Activities in and Against (Nicaragua v. U.S.) (Merits), 1986 I.C.J. Rep.
6, ¶¶ 132–134 (dissenting opinion of Judge Schwebel); Kasikili/Seduku Island (Botswana v.
Namibia) (Merits), 1999 I.C.J. Rep. 1045, 1118–19 (separate opinion of Judge Oda); Mari-
time Delimitation and Territorial Questions (Qatar v. Bahrain) (Merits), 2001 I.C.J. REP. 40,
275 (dissenting opinion of Judge Torres Bernárdez); Pulp Mills on the River Uruguay (Ar-
gentina v. Uruguay) (Merits), 2010 I.C.J. REP. 14, 108. In the first of these cases, the P.C.I.J.
in the Oscar Chinn case declined—against significant dissent—to appoint an expert to assist
in determining whether Belgium had established a de facto monopoly in the Belgian Congo,
in violation of its international obligations to the United Kingdom. See Oscar Chinn (U.K. v.
Belgium), Judgment, P.C.I.J., ser. A/B, No. 63, at 88 (1934) (reasoning that the details of the
economic conditions related to the claims could not change the legal conclusion that Belgium
had not violated its international obligations); id. at 108 (individual opinion of Judge
Anzilotti) (“[T]his necessitates the appraisement of a number of technical considerations
which only experts in fluvial navigation and in the economic conditions of the Congo could
adequately present to the Court.”); see also SANDIFER, EVIDENCE BEFORE INTERNATIONAL
TRIBUNALS, supra note 10, at 334–35 (1975) (criticizing the Oscar Chinn majority’s refusal to
appoint an expert); Thomas M. Franck, Fact-Finding in the I.C.J., in FACT-FINDING BEFORE
INTERNATIONAL TRIBUNALS 21, 21 (1992) (faulting the Court for resolving “key factual is-
sues” in Temple of Preah Vihear, the Nicaragua Case, and the Advisory Opinion in Western
Sahara “without leaving The Hague to take testimony, appointing masters, or otherwise fa-
miliarizing itself with the sights, sounds and smells of the place in which the facts were
embedded.”).
for refusing to consider expert evidence. The first case, Maritime Delimi-
tation and Territorial Questions Between Qatar and Bahrain, centered on
whether the landmass Fasht Al Azm and Sitrah Island were joined, an
issue that affected the extent of each country’s maritime areas.144 Qatar
argued that a navigable channel had always separated the two until
Bahrain filled it in 1982, and the parties submitted competing expert re-
ports on the existence of a permanently navigable channel before that
time.145 The Court viewed the expert evidence as too difficult to evaluate
and based its decision on other issues without reaching a conclusion on the
pre-1982 existence of a channel.146 Others have noted that the Court-de-
termined “maritime” boundary inadvertently crosses over dry land, a situ-
ation that would have likely not occurred had the Court relied on expert
assistance.147
In its 1997 Gabèı́kovo-Nagymaros Project decision, the ICJ essentially
refused to consider the parties’ expert evidence148 and this decision has
been cited as an example of the ICJ’s inability or unwillingness to deal
with technical evidence.149 At issue in Gabèı́kovo-Nagymaros was the
agreement by Hungary and Slovakia in a 1977 treaty to jointly build a
system of hydroelectric-generating locks on the Danube River.150 The dis-
pute arose after Hungary, facing public criticism of the financial benefits
and environmental impact of the project, halted construction of the pro-
ject, citing a “state of ecological necessity.”151
At the ICJ, both parties submitted expert reports on the ecological
effects of the locks, with Hungary arguing that they would cause eutrophi-
cation in the reservoir, a build-up of silt within the dredged channel, and
negative effects on the plants and animals that would be separated from
the main river channel.152 In the Court’s view, however, it was “not neces-
sary . . . to determine which of those points of view is scientifically better
founded” because the environmental risks were at most long-term and un-
certain, abandonment of the project was not the only mitigating measure
Hungary could have taken, and Hungary had helped create the situation
when it concluded the 1977 treaty.153
A similar pattern was repeated in 2010, in the Pulp Mills on the River
Uruguay case. The case readily lent itself to technical evidence, as it cen-
144. Maritime Delimitation and Territorial Questions Between Qatar and Bahrain
(Qatar v. Bahrain), Judgment, 2001 I.C.J. Rep. 40, ¶¶ 217–218 (2001).
145. See id. ¶ 189.
146. See id. ¶¶ 190, 218.
147. See, e.g., Riddell, supra note 28, at 243.
148. See Gabèı́kovo-Nagymaros Project (Hungary v. Slovakia), Judgment, 1997 I.C.J.
Rep. 7, 42 (1997).
149. ANNA RIDDELL & BRENDAN PLANT, EVIDENCE BEFORE THE INTERNATIONAL
COURT OF JUSTICE 348 (2009).
150. Gabèı́kovo-Nagymaros, Judgment, 1997 I.C.J. Rep. at 23.
151. Id. at 35.
152. Id.
153. Id. at 42, 55, 57.
tered on Argentina’s claim that a Uruguayan plan to build pulp mills on

the River Uruguay, the boundary between Uruguay and Argentina, would
significantly pollute the river, in breach of a 1975 treaty between the two
countries.154 Even though the parties produced “a vast amount of factual
and scientific material,”155 the Court did not discuss the merits of the ex-
tensive expert evidence and expressly stated that it would not address the
reliability of that evidence or the weight accorded to it.156 In dissent,
Judges Al-Khasawneh and Simma lamented the Court’s approach, reason-
ing that “[t]he Court on its own is not in a position adequately to assess
and weigh complex scientific evidence of the type presented by the Par-
ties.”157 Instead, “[t]he adjudication of disputes in which the assessment
of scientific questions by experts is indispensable, as is the case here, re-
quires an interweaving of legal process with knowledge and expertise that
can only be drawn from experts properly trained to evaluate the increas-
ingly complex nature of the facts put before the Court.”158
C. Use of Expert Evidence as a Veneer to Enhance Legitimacy

Given the amount of time and paper devoted to expert evidence and
the potential influence that it may have on the outcome of international
disputes, we would expect that the written decisions resolving those dis-
putes would at the very least explain whether and how the decision-mak-
ers relied on expert evidence.159 But too many decisions do not do that.
Failure to record in written decisions the extent of reliance on expert evi-
dence compounds the potential for undue reliance on or unreasoned rejec-
tion of expert evidence. As a dissent in the 1949 ICJ Corfu Channel Case
154. Pulp Mills on the River Uruguay (Arg. V. Uru.), Judgment, 2010 I.C.J. Rep. 14, ¶¶
1, 22 (Apr. 20).
155. Id. ¶ 165; see also Bruno Simma, The International Court of Justice and Scientific
Expertise, 106 AM. SOC’Y INT’L L. 230, 230 (2012) (noting that both parties in the Pulp Mills
case produced “large amounts of scientific expert evidence, both in their written pleadings
and in the oral hearings”).
156. Pulp Mills on the River Uruguay, 2010 I.C.J. Rep. 14, ¶¶ 166, 168 (“[T]he Court
does not find it necessary in order to adjudicate the present case to enter into a general
discussion on the relative merits, reliability and authority of the documents and studies pre-
pared by the experts and consultants of the Parties.”); see also id. ¶ 236 (“[I]n assessing the
probative value of the evidence placed before it, the Court will principally weigh and evalu-
ate the data, rather than the conflicting interpretations given to it by the Parties or their
experts and consultants.”); FOSTER, supra note 3, at 93 (noting that the ICJ, in Pulp Mills,
“did not find it necessary to enter into a general discussion on their merits, reliability[,] and
authority.”).
157. Pulp Mills on the River Uruguay (Arg. v. Uru.), 2010 I.C.J. Rep. 108, ¶ 4 (Apr. 20)
(Al-Khasawneh, J. and Simma, J., joint dissenting opinion).
158. Id. ¶ 3.
159. Foster, supra note 45, at 161 (“We can expect judgments to be written . . . in terms
reflecting how the [ICJ] has made use of expert opinion. Transparency in the reasoning of
judgments will help them to stand up to scrutiny, including where expert opinion has fed into
the more legal and interpretive tasks required of the Court.”); Pauwelyn, supra note 77, at
189 (“[I]t is equally important that the judge when weighing the evidence sufficiently explains
his or her thought process and engages with the details of each study.”).
reasoned, “[a]ccording to a quite general rule of procedure, the Court is

not bound by the opinion of experts [and] [t]he Court may reject or accept
it[,] but it must always give sufficient reasons.”160
Others have argued that decision makers often marshal enormous
amounts of expert evidence as part of the decision-making process, or in
their written decisions, as a way to enhance their perceived legitimacy,161
even in venues, such as the WTO, that routinely seek input from ex-
perts.162 From this perspective, a decision maker may cite to expert evi-
dence to demonstrate that it has left no stone unturned, without engaging
with the evidence to better understand the substance of the issues in dis-
pute or to assess its quality.
Expert evidence that complicates rather than quickly confirms other
aspects of a decision maker’s analysis may create incentives to ignore the
evidence in the written decision, but the reticence to make meaningful use
160. Corfu Channel (U.K. v. Alb.), 1949 I.C.J. Rep. 4, 253 (Dec. 15) (Eèer, J., dissent-
ing opinion).
161. See Bonneuil & Levidow, supra note 49, at 97 (criticizing WTO decisions under the
SPS Agreement for failing to engage with scientific evidence and framing that evidence” in a
way that allows WTO judges to avoid any explicit engagement with scientific knowledge,” so
that they can use that evidence as an legitimizing “imprimatur” for “its own judgements on
substantive scientific issues”); see also Lazareff, supra note 111, at 17 (“Assessing damages is
the parent pauvre of arbitration, the neglected aspect. It is almost in the context of arbitra-
tion, the midnight clause of a contract, and it is very distressing to read in so many awards
that ‘the Tribunal, having at its disposal all the elements of the case, orders A to pay B US $140
million.’ ”); Simma, supra note 155, at 230 (noting that the parties in the Pulp Mills case, see
supra note 156, “produced large amounts of scientific expert evidence” and explaining that as
an ICJ judge in the case the author, when “confronted with the mass of such highly complex
technical material” did “not feel[ ] capable of drawing the necessary legal conclusions”).
162. See FOSTER, supra note 3, at 133 (“[E]ven in the WTO, the full impact of expert
evidence is not always apparent on the face of the report. For example in United States –
Import Prohibition of Certain Shrimp and Shrimp Products, the Appellate Body’s final dispo-
sal of the case was undoubtedly informed by the experts’ advice that sea turtles faced differ-
ent problems in different locations, but in reaching this decision the Appellate Body did not
directly rely on the panel’s consultation of scientific experts.”); cf. Walker, supra note 62, at
300–02 (“The [EC —]Hormones Panel took the worst approach [in evaluating whether a
measure was “based on” a risk assessment in accordance with the SPS Agreement]: (1) pre-
tend not to pass judgment on the merits of past scientific reports while implicitly agreeing
with their conclusions, and (2) formally preclude rebuttal explanation in the WTO proceed-
ing by members who disagree with those scientific reports, saying that WTO panels do not
themselves conduct risk assessments.” (citation omitted)); id. at 310
[T]he [EC — Hormones] Panel was attempting an impossible task—making deter-

minations about risk assessment documents without actually understanding or tak-
ing a position on risk assessment issues. . . . The Hormones Panel should have used
all of the relevant scientific information made available to it by the parties and by
its own appointed experts, but should have used it for the limited purposes of find-
ing whether there was any reasonable scientific basis for the risk determinations of
the European Communities.
(footnote omitted).
of expert evidence can be so strong that decision-makers will not rely on it

even when it appears to support their conclusions.163
IV. MORE ROBUST ASSESSMENTS OF EXPERT RELIABILITY

How can legal decision makers address these challenges and enhance
their handling of expert evidence? Part II of this Article outlined recent
proposals to improve the use of expert evidence by international courts
and tribunals through procedural and transparency mechanisms, including
enhanced conflict-of-interest disclosure requirements and hearings that
emphasize joint evaluations of competing experts. While these mecha-
nisms may prove valuable, they are not enough. As others have recog-
nized, decision makers cannot engage passively with complexity,
particularly when it comes to expert evidence.164 From the perspective of
ICJ Judges Al-Khasawneh and Simma, “the traditional methods of evalu-
ating evidence are deficient in assessing the relevance of such complex,
technical and scientific facts.”165 Decision makers must recognize the
“strengths and weaknesses of expert opinions,” meaningfully distinguish
among competing opinions and methodologies, and understand whether
an expert’s methodology can be reliably applied to the specific circum-
stances of the case at hand.166 Procedures and transparency alone cannot
163. See Jürgen Kurtz, Science as a Common Proxy for Rational Regulation Across
International Trade and Investment Law, in SCIENCE AND TECHNOLOGY IN INTERNATIONAL
ECONOMIC LAW: BALANCING COMPETING INTERESTS 134, 146 (Bryan Mercurio & Kuei-Jung
Ni eds., 2014) (noting that the Methanex tribunal’s “careful and extensive assessment of the
environmental justifications for the MTBE ban” based on “significant attention to extensive
witness testimony” “receives no attention whatsoever in the Tribunal’s legal analysis of fair
and equitable treatment” but “is instead quarantined in the factual findings of the award,”
even though “this review . . . provides clear and compelling evidence, as the Tribunal itself
determines, of a clearly rational and science-based approach to risk regulation”).
164. See MARBOE, supra note 64, at 4 (reasoning, in the context of damages calculations
in investment disputes, that “[i]n order to assess the figures submitted by experts, a basic
understanding of the methods and a willingness to consider them is of vital importance” to
avoid delegating “the decision about the amount of compensation or damages” to the ex-
perts); RIPINSKY & WILLIAMS, supra note 43, at 180 (noting in the valuation context that
“arbitrators may need to make diligent efforts to understand economic and valuation issues
on their own” to be able “to make use of experts’ reports and at the same time exercise their
decision-making power”); Foster, supra note 45, at 144 (“An international court must be
ready to determine objective compliance with legal tests in order to enforce the balance of
interests they represent. In order to achieve this, the Court [ICJ] will need to engage in the
science, rather than to be viewed as shying away from doing so.”); id. (“The Court [ICJ] will
need to ensure it has the capacity to achieve the sufficiently reliable insights into the science
necessary for a sound resolution of both disputes, taking into account also the interest s of
the wide international community.”).
165. Pulp Mills on the River Uruguay (Arg. v. Uru.), 2010 I.C.J. Rep. 108, ¶ 3 (Apr. 20)
(Al-Khasawneh, J. and Simma, J., joint dissenting opinion).
166. See RIPINSKY & WILLIAMS, supra note 43, at 190–91
[L]awyers need to have an idea about the basic workings of valuation techniques in
order to recognize the strengths and weaknesses of expert opinions, be able to
discern the ‘drivers’ of valuation, understand why different valuation techniques
may yield widely divergent results and, indeed, why even the same valuation
ensure that expert evidence is reliable and that decision makers reliably
understand and apply that evidence in crafting legal decisions. As Lucy
Reed has reasoned in the context of investor-state arbitration, decision-
makers “can do a better job” in fact-finding “by following basic practices
to control and mitigate complexities.”167
When it comes to mitigating the complexities of expert evidence, the
focus should be on the reliability of that evidence: the reliability of the
methodology and the reliability of that methodology’s application to the
facts of the case. The core of this Article’s proposal, then, is that interna-
tional adjudicators should adopt an analytical framework for determining
whether an expert’s testimony is reliable, apply that framework to each
expert, clearly set out in the written decision the framework and how it
was applied, and weigh the evidence in accordance with its assessed
reliability.
A. Applying a Reliability Framework
The core of this Article’s proposal is an illustrative and flexible check-

list for evaluating the reliability of expert evidence. In other fields, as
“[k]now-how and sophistication have increased remarkably,”168 decision-
makers have turned to checklists to “correctly” and “reliably” deliver the
benefits of that know-how and sophistication.169 Engineers, doctors, and
pilots—to name just a few professions—use them to remind themselves of
the “minimum necessary steps” and to “establish[ ] a higher standard of
baseline performance.”170 Checklists, then, can “help even experienced
professionals” to “slow down and take account of all relevant factors”171
and to ensure that decision-makers have received all information critical
to their decision.172 And because checklists also set out the analytical
steps for all to see, they can lead to more transparent decision-making.173
method applied by different experts can lead to very different results. This knowl-
edge is necessary for an informed and considered decision.
See also Walker, supra note 62, at 169 (“[T]he SPS Agreement’s approach of using scientific
research as a neutral arbiter over the legitimacy of sanitary and phytosanitary measures de-
pends upon the WTO’s being able to distinguish legitimate or sound science from an inade-
quate risk assessment.”).
167. Lucy Reed, Confronting Complexities in Fact-Finding and the Nature of Investor-
State Arbitration, in AM. SOC’Y INT’L L., PROCEEDINGS OF THE 106TH ANNUAL MEETING
233, 233 (2013).
168. ATUL GAWANDE, THE CHECKLIST MANIFESTO 11 (2009).
169. Id. at 13.
170. Id. at 39.
171. Reed, supra note 49, at 14 (“One thing we see in heuristics studies is the value of
checklists. Checklists help even experienced professionals—for example, pilots and sur-
geons—to slow down and take account of all relevant factors.”).
172. GAWANDE, supra note 168, at 79, 99, 101.
173. Id. at 79, 99, 101.
When considering the “evidential value of any evidence produced . . .

under all the circumstances” of a highly complex international dispute,174
an expert reliability checklist could perform similar functions. The result
of the checklist would be a clearer analytical roadmap—clearer to the dis-
puting parties, the public, and future decision makers—that a tribunal can
use to systematically weigh expert testimony and ultimately the merits of
the parties’ arguments.175 More reliable and relevant testimony would be
entitled, as it is now, to greater weight than less reliable and relevant
testimony.176
At the same time, any proposal for evidence-taking in international
law should rely on insights developed within international dispute resolu-
tion systems and avoid disrupting the essential elements of those systems.
Because international law recognizes no hard-and-fast rules for assessing
the reliability of expert evidence, leaving adjudicators free to assess and
weigh evidence as they see fit,177 this proposal does not seek to limit the
flexibility or discretion inherent in international adjudication. A checklist
174. Shufeldt (U.S. v. Guat.), 2 R.I.A.A. 1079, 1083 (1930) (“[I]t is clear that interna-
tional courts are by no means as strict as municipal courts and can not [sic] be bound by
municipal rules in the receipt and admission of evidence. The evidential value of any evi-
dence produced is for the international tribunal to decide under all the circumstances of the
case.”).
175. See Pauwelyn, supra note 77, at 185 (suggesting that adjudicators adhere to,
“openly state,” and “enforce” a set of “best practices” based on European competition law in
evaluating quantitative economic data).
176. See, e.g., Corfu Channel (U.K. v. Alb.), Judgment, 1949 I.C.J. Rep. 4, 21 (Dec. 15)
(stating that the Court “cannot fail to give great weight to the opinion of the [e]xperts who
examined [the evidence] in a manner giving every guarantee of correct and impartial infor-
mation.”); U.S. DEP’T OF STATE, supra note 86, at 428–29 (noting that while “[i]n an arbitra-
tion between States it is of far greater interest than in purely juridical proceedings to draw
forth all evidence, whether direct or indirect, which may serve to give full light,” the arbitra-
tor would take into account the circumstances of the expert testimony when weighing the
evidence); see also Michael Straus, The Practice of the Iran-U.S. Claims Tribunal in Receiving
Evidence from Parties and from Experts, 3 J. INT’L ARB. 57, 58 (1986) (“[A]ll testimony is
ultimately subject to the arbitrators’ judgment as to its probative value based on such factors
as credibility and relevance.”); JEFF WAINCYMER, PROCEDURE AND EVIDENCE IN INTERNA-
TIONAL ARBITRATION 943 (2012) (“In most cases, the [arbitral] tribunal will simply allow the
parties to present the experts of their choice and allow submissions as to relative expertise to
go to questions of weight.”).
177. See GARY BORN, INTERNATIONAL ARBITRATION: CASES AND MATERIALS 767
(2015) (“[A]rbitrators in interstate arbitrations enjoy broad discretion to make evidentiary
decisions on subjects such as admissibility, weight, and credibility of evidence, usually with-
out reference to municipal evidentiary rules.”); FOSTER, supra note 3, at 178 (noting that
international law recognizes no “rules requiring that weight will only be placed on scientific
evidence that has been subject to peer review and publication”); DURWARD V. SANDIFER,
EVIDENCE BEFORE INTERNATIONAL TRIBUNALS 2 (1939) (“[I]nternational tribunals are . . .
intolerant of any restrictive rules of evidence which might tend to confine the scope of a
search after those facts.”); J. L. SIMPSON & HAZEL FOX, INTERNATIONAL ARBITRATION: LAW
AND PRACTICE 192 (1959) (“In international law there are no general rules requiring the
exclusion of categories of evidence.”); see also WHITE, supra note 9, at 7 (quoting Sandifer’s
assertion with approval).
is not a formula,178 and decision makers would of course remain free to

tailor the checklist—add, subtract, or modify elements—to fit particular
circumstances of a dispute. Procedural and evidentiary procedures would
thus remain flexible to adapt to the particular circumstances of each
case.179
]B. A Proposed Checklist to Evaluate Expert Reliability

This Article sets out six suggested framing questions—drawn from in-
ternational law decisions, practice, and commentary—that might usefully
serve as a starting point for evaluating the reliability of expert evidence.
These questions emphasize the reliability of the process by which an ex-
pert has developed and presented evidence, not on the validity of the out-
come of that process per se. The proposed checklist seeks to focus the
inquiry on proxies that are most accessible to lawyers and judges—that is,
most like the ways in which they examine and weigh other types of evi-
dence—and does not demand that a decision maker become the arbiter of
competing claims of truth in technical fields where she lacks the necessary
competence and training.180
Importantly, when presented with expert evidence, a decision maker
ought to record in the written decision whether, how, and why it has relied
on that evidence. A record of the question presented—including applica-
ble legal and evidentiary rules—and of the extent to which a decision
maker relied on expert evidence in answering that question is essential to
avoiding any doubt about the relative roles of expert and legal decision
maker.181 Without expressly connecting the question presented, the ex-
178. GAWANDE, supra note 168, at 167.

179. Because “[e]ach arbitration presents its own peculiar problems,” “[r]ules that may
function well enough in one arbitration may lead to a complete breakdown when introduced
in another arbitration possessing totally different characteristics.” KENNETH S. CARLSTON,
THE PROCESS OF INTERNATIONAL ARBITRATION 30–31 (1946). “[P]rocedural rules should”
thus “be carefully adapted to the requirements of each arbitration as it arises so that it may
be consummated speedily, economically and justly.” Id. at 4. “[T]he type of the procedure
can always be adjusted to the complexity and volume of the litigation to be submitted for
decision.” Id. at 260. Domestic evidentiary rules also do not prescribe strict rules that courts
must follow when examining the reliability of experts. The U.S. Supreme Court, for example,
has recognized that “the factors identified in Daubert [v. Merrell Dow Pharmaceuticals, Inc.]
may or may not be pertinent in assessing reliability, depending on the nature of the issue, the
expert’s particular expertise, and the subject of his testimony.” See Kumho Tire Co. v. Car-
michael, 526 U.S. 137, 150 (1999).
180. See Marcos A. Orellana, The Role of Science in Investment Arbitrations Concern-
ing Public Health and the Environment, 17 Y.B. INT’L ENVTL. L. 48, 49 (2006) (“[E]mphasis
on the process by which the scientific enterprise has been conducted, rather than on absolute
scientific validities, may prevent arbitrators from becoming ensnared in deciding over the
truth of scientific claims, especially in situations of scientific uncertainty.”).
181. See Francesca Romanin Jacur, Remarks on the Role of Ex Curia Scientific Experts
in International Environmental Disputes, in INTERNATIONAL COURTS AND THE DEVELOP-
MENT OF INTERNATIONAL LAW 441, 452 (Nerina Boschiero et al. eds., 2013) (“Transparency
should also be reflected in the final judgment or award. The reasoning of the judiciary should
be spelled out clearly, showing how certain conclusions were reached and the contribution of
the experts.”).
pert evidence, and the decision, a court or tribunal may leave itself open to
criticisms, including those outlined in Part III.
1. Will expert evidence usefully aid in answering the question

presented?
The threshold question in determining the weight to accord expert tes-

timony should be whether the testimony is relevant to the question
presented and will aid in the decision-making process. If the issue can be
decided on the basis of other evidence—for example, documentary evi-
dence and fact witness testimony—or if the question requires only a legal
determination, a decision maker may have no need to rely on an expert’s
testimony.182 Asking this threshold question is important because, al-
though expert testimony can assist in clarifying an adjudicator’s decision, it
can also confuse a question that would have been otherwise easily and
transparently decided without the expert’s input. Answering the ques-
tion—in writing—also avoids any doubt as to whether, how, and why a
decision-maker has relied on expert evidence in reaching its decision.
Today, arbitrators and courts are broadly empowered to consider and
rely on expert testimony as they see fit.183 But of course just because a
tribunal can rely on expert testimony does not mean that it should. The
power to consider expert testimony must include the power to afford only
the weight that the testimony deserves. This is a natural inquiry in civil
law jurisdictions and in international adjudicatory systems where the tribu-
nal—not the parties—decides whether to appoint experts.
What is the technical question presented?—As simple as it may seem, it
is essential to begin with a clear sense of the question that is presented,
based on the applicable legal and evidentiary rules, and the factual circum-
stances. What was the value of an investment at the time of its expropria-
182. See Pulp Mills on the River Uruguay (Arg. v. Uru.), 2010 I.C.J. Rep. 121, ¶ 11
(Apr. 20) (Keith, J., concurring opinion) (“[R]esponsibility of making decisions on the mat-
ters of scientific dispute arises only if the matters require decision in the course of the Court
determining whether or not Argentina had made out its claim.”); Factory at Chorzów (Ger. v.
Pol.), Judgment, 1928 P.C.I.J. (ser. A) No. 17, 49, 51 (Sept. 13) (noting that the Court was not
“satisfied with the data for assessment supplied by the Parties” with respect to compensation
owed to the German government and providing for the appointment of experts “to obtain
further enlightenment in the matter”); Starrett Housing Corp. v. Iran, Final Award, 16 Iran-
U.S. Cl. Trib. Rep. 112, ¶ 264 (1987) (noting that, although it is not always necessary to
appoint valuation experts where “there is sufficiently clear evidence on which to base a deci-
sion on compensation,” tribunals may look to expert evidence “to obtain any technical infor-
mation that might guide it in the search for the truth” (partially quoting Corfu Channel (U.K.
v. Alb.), Merits, 1949 I.C.J. Rep. 4, 20 (Dec. 15))); see also Dave, supra note 28, at 814 (“It is
important to bear in mind that if direct evidence is available and is acceptable, it is hardly
necessary to consider expert opinion.”); Freyer, supra note 44, at 443 (“The ICC reminds us
that expert evidence is not a requirement of an arbitration, and should only be introduced to
the extent necessary to inform the tribunal on a key issue in dispute.”).
183. See supra notes 87 & 93 and accompanying text.
tion184 or on the date of the award?185 What evidence exists that eating
meat from farm animals that were treated with hormones for growth-pro-
motion increases the risk of cancer in humans?186 What is the maritime
boundary between the continental and fisheries zones of two countries?187
Has construction of man-made islands negatively affected surrounding
coral reef systems?188 Knowing the question to be answered in precise
and clear terms provides an essential basis for determining what evidence
would be helpful in answering it.189 This is second nature when evaluating
tribunal-appointed experts190 but no less important when evaluating
party-appointed experts.
In formulating potential questions for an expert, decision makers
should always acknowledge the extent to which a legal rule has informed
the question scope, lest they open themselves up to criticisms that they are
unduly allowing expert evidence to influence those rules191 or requiring
184. See, e.g., Ebrahimi v. Iran, Award, 30 Iran-U.S. Cl. Trib. Rep. 170, 182–83 (1994);
Starrett Housing Corp. v. Iran, 16 Iran-U.S. Cl. Trib. ¶ 4.
185. See, e.g., Siemens A.G. v. Argentine Republic, ICSID Case No. ARB/02/08,
Award, ¶¶ 352–353 (Feb. 6, 2007); Ger. V. Pol., 1928 P.C.I.J. (ser. A) No. 17, at 47, 51–54
(Sept. 13).
186. See US — Continued Suspension, supra note 111, ¶ 200 (examining whether scien-
tific evidence “focus[ed] on” and “address[ed] . . . the carcinogenic or genotoxic potential of
the residues of those hormones found in meat derived from cattle to which the hormones had
been administered for growth promotion purposes – as required by paragraph 4 of Annex A
of the SPS Agreement”).
187. E.g., Delimitation of the Boundary in the Gulf of Maine Area, Judgment, 1984
I.C.J. Rep. 263 (“What is the course of the single maritime boundary that divides the conti-
nental shelf and fisheries zones of Canada and the United States of America?”).
188. See Philippines v. China, Award, PCA Case No. 2013-19 (2016), ¶ 84 (2015).
189. See Jacur, supra note 181, at 451 (“The adjudicator should properly identify the
extent of the scientific matter at stake and then accordingly carefully pose the questions to
the [tribunal-appointed] experts.”); Pauwelyn, supra note 77, at 186
The first step [in evaluating quantitative economic data] is precisely formulating
the relevant question, that is, a question whose answer will be relevant in deciding
the applicable legal criterion (e.g., what was the effect of US cotton subsidies pro-
vided in year(s) X on world prices of cotton so we can figure out whether these
subsidies caused ‘serious prejudice’ in the sense of SCM [the WTO Agreement on
Subsidies and Countervailing Measures] Article 6.3(c)).
190. See, e.g., Corfu Channel (U.K. v. Alb.), Judgment, 1949 I.C.J. Rep. 4, 142–69 (Dec.
15) (posing a series of questions to the tribunal-appointed experts); see also Lukasz
Gruszczynski, The Role of Experts in Environmental and Health-Related Trade Disputes in
the WTO: Deconstructing Decision-Making Processes, in THE ROLE OF “EXPERTS” IN INTER-
NATIONAL AND EUROPEAN DECISION-MAKING PROCESSES: ADVISORS, DECISION MAKERS
OR IRRELEVANT ACTORS 216, 221–22 (M. Ambrus et al. eds., 2014) (explaining that the pro-
cess of consulting experts in WTO disputes under the SPS Agreement “starts with the deter-
mination of the necessary fields of expertise and is followed by the identification of potential
candidates” and the “draft[ing] [of] questions relating to the specific scientific or technical
aspects of a dispute”).
191. See Howse, supra note 32, at 2345–46 (faulting the WTO Panel in Australia —
Salmon for “never plac[ing] the issues [for expert evaluation] in the context of the legal
meaning of the SPS Agreement provisions on risk assessment and scientific evidence” and
for “fail[ing] to articulate how it intended to use or weigh the scientific evidence in deciding
experts to answer questions outside their technical competence.192 When

formulating a question for damages experts, for example, a decision maker
could, depending on the case, specify the exact contours of the enterprise
or other asset to be valued, the relevant valuation date or dates, and
whether certain events or facts should be considered or not considered
under the applicable legal rules.193
Does non-expert evidence or a legal rule obviate the need for expert
evidence? With a question in hand, the decision maker should then deter-
mine whether available non-expert evidence is sufficient to answer that
question or whether the question is purely legal in nature, so that expert
evidence is not useful to the decision-making process. In boundary dis-
putes between neighboring countries, for example, decision-makers have
often turned to expert evidence only after acknowledging that other sorts
of evidence are not enough to draw a definitive boundary.194 In disputes
centering on the design, construction, or operation of industrial facilities
or extractive concessions, decision-makers have often determined that ex-
pert evidence is required.195 Similarly, when it comes to calculating the
amount of damages owed, decision makers often have found documentary
evidence alone insufficient to make an appropriate assessment.196 Experts
the question of adequacy”); see also Appellate Body Report, Australia — Salmon, ¶ 6, WTO
Doc. WT/DS18/AB/R, (adopted Oct. 20, 1998).
192. See Howse, supra note 32, at 2347 (“The scientists called upon in Salmon were
placed in a virtually impossible position: they were asked to make a purely technical/scientific
judgment about the adequacy of risk assessment as a regulatory tool.”).
193. See RIPINSKY & WILLIAMS, supra note 43, at 179 (“Tribunals can improve compa-
rability of expert reports by providing guidance on the essential parameters of the task to be
performed, for example on: exact object of valuation; relevant valuation date or dates; and
whether the impact of certain events subsequent to the valuation date should be accounted
for.”).
194. See, e.g., Honduras Borders (Guat. v. Hond.), 2 R.I.A.A. 1307, 1353 (1933) (noting
that the Court had commissioned a report from party-appointed experts in light of the inade-
quacy of existing topographical data); MOORE, supra note 25, at 297–98
No geographical map was annexed to the treaty . . . and in our opinion none which
can be adopted as a sure and complete proof of the intentions of the negotiators of
the treaty. . . . [D]uring the arbitration proceedings no map was produced which
was recognized as being entirely accurate by both parties. . . . [The parties] dis-
cussed much about the importance and accuracy of their maps, but . . . these dis-
cussions did not lead to any decided conclusion as to the value to be given to one
of these maps more than to the other as regards the various features of the
frontier.
195. Harza v. Iran, Interlocutory Award, 2 Iran-U.S. Cl. Trib. Rep. 68, 71 (1983)
[T]he Tribunal cannot decide the liability issues raised by the various alleged de-
fects of site selection, engineers’ drawings, and actual construction of major dams
and related irrigation networks without a proper technical description of the al-
leged defects and of the causes from which they arose. This will require opinions
to be sought at least from two experts.
196. In the Lighthouses Arbitration, for example, the Permanent Court of Arbitration
determined that expert testimony was necessary because “the documents so far lodged by the
parties” were not “sufficient to determine” or “capable of furnishing an exact assessment” of
have also often been deemed useful on questions related to the business
and commercial practices in a particular industry.197
By contrast, in a series of cases involving the delimitation of maritime
boundaries, the ICJ determined that a legal rule—which generally enables
a country to claim a continental shelf up to as far as 200 miles from its
coast—meant that it need not weigh expert evidence on the specific geo-
logical characteristics offshore of each disputing country.198 The ICJ in-
stead reasoned that its function is “to make use of geology only so far as
required for the application of international law.”199
Does the expert (or would an expert) have specialized knowledge be-
yond that of the decision-maker? Evidence offered as “expert” will in gen-
eral aid in the decision-making process only if the witness has specialized
knowledge, skill, or training that complements and goes beyond the com-
petence and training of the adjudicator in answering the specific question
presented.200 Two of the first modern cases to use experts extensively in
their decision-making processes illustrate the importance of this question.
First, in the Manica Plateau Arbitration, a boundary dispute, the arbitrator
explained that he had sought out an expert “specially qualified in ques-
the total costs and expenditures of the concession at issue. See Affaire relative à la conces-
sion des phares de l’Empire ottoman, (Fr. v. Greece) 12 R.I.A.A. 155, 228 (Perm. Ct. Arb.
1956) (unofficial translation by author).
197. Harza v. Iran, 2 Iran-U.S. Cl. Trib. at 71 (“[I]t appears from the pleadings and the
substantiating evidence submitted by both Parties that the reasons for the disagreement be-
tween the Parties need to be clarified to the Tribunal by an expert familiar with the business
practices and administration of important consulting engineering contracts.”).
198. See, e.g., Continental Shelf (Libya v. Malta), Judgment, 1985 I.C.J. REP. 13, 36
(June 3); Continental Shelf (Tunis. v. Libya), Judgment, 1982 I.C.J. REP. 18, 41, 53, 58 (Feb.
24).
199. See Tunis. v. Libya, 1982 I.C.J. at 54; see also Tullio Scovazzi, Between Law and
Science: Some Considerations Inspired by the Whaling in the Antarctic Judgment, 14 QUES-
TIONS INT’L L.: ZOOM-IN 13, 21 (2015) (noting that, in these maritime boundary cases, “where
it is possible to simply excessively complex questions for the sake of legal clarity, one cannot
but agree with the approach taken by the I.C.J. in disregarding scientific discussions and
experts’ elaborations – be they appointed by the parties or by the Court”).
200. See, e.g., Prosecutor v. Peris̆ić, Case No. IT-04-81-T, Decision on Defense Motion
to Exclude Expert Reports of Robert Donia, ¶ 6 (Int’l Crim. Trib. For the Former Yugoslavia
Oct. 27, 2008) (noting that the jurisprudence of the International Criminal Tribunal for the
former Yugoslavia defines an expert as “a person whom by virtue of some specialised knowl-
edge, skill or training can assist the trier of fact to understand or determine an issue in dis-
pute”); I.V.E.M. Claim, 22 Int’l L. Rep. 875 (1955) (noting that, according to the Franco-
Italian Conciliation Commission, international adjudication often leads to “inquiries and as-
sessments which presuppose technical knowledge denied to [their own members]”);
CHARLES N. BROWER & JASON D. BRUESCHKE, THE IRAN-UNITED STATES CLAIMS TRIBU-
NAL 198 (1998) (“In many international arbitrations, including the Tribunal, the arbitrators
are not specialists in the underlying subject matter. Given the broad spectrum of its cases
and the complexity of issues presented for adjudication, it therefore not surprising that the
use of informed analysis of experts has been invaluable to the Tribunal’s understanding and
resolution of . . . issues.” (citations omitted)); Scovazzi, supra note 199, at 16 (“Nor can the
members of a court, who are experts in law, be supposed to have a universal knowledge, in
order to reach by themselves conclusions that require scientific and technical expertise. Here
scientific and technical experts have a role to play.”).
tions of geography and topography” because the arguments of the parties

were not within his training or competence but were instead “of an essen-
tially technical character.”201 As with many boundary disputes, Manica
Plateau centered on “technical” questions of geography, topography, and
hydrology, and how to account for the myriad, overlapping features on the
ground—rivers, watersheds, ravines, valleys, slopes, mountains, and
plateaus—that defy simple legal analysis.202
Second, in the Corfu Channel case, the ICJ asked naval officers to
opine on whether a mine-laying operation in a strait that had damaged a
British ship could have been conducted without Albanian authorities on
land seeing or hearing the operation, how recently the mines had been
laid, and whether the mines were floating or moored.203 Without prior
experience with such operations, it would not be possible to assess the
type of operation necessary, its visual footprint, or the level of noise it
would create, nor would a layperson have the knowledge or experience to
carry out a re-creation and examination of a naval mine-laying
operation.204
2. Does the witness have the requisite experience and training in the
chosen methodology?
If expert evidence will aid the decision-making process, a decision-
maker should next determine whether the expert has the necessary cre-
dentials and experience in the relevant field to act as a reliable author-
ity.205 Evaluating the appropriate range of expertise can help to neutralize
strategies designed to obfuscate genuine issues, and provide decision-mak-
201. Manica Plateau Arbitration (Gr. Brit. v. Port.), 28 R.I.A.A. 283, 308 (1897). The
focus of this article is on non-legal experts, although courts and tribunals also often consider
testimony on municipal and international law. But See Andrew Newcombe, The Strange
Case of Expert Legal Opinions in Investment Treaty Arbitrations, KLUWER ARBITRATION
BLOG (Mar. 18, 2010), http://kluwerarbitrationblog.com/2010/03/18/the-strange-case-of-ex-
pert-legal-opinions-in-investment-treaty-arbitrations.
202. See 28 R.I.A.A. 283, 308 (1897); see also I.V.E.M. Claim, Final Decision, V
RECUEIL DES DÉCISIONS DE LA COMMISSION DE CONCILIATION 153, 177 (Mar. 7, 1955) (ex-
plaining that experts were appointed because the issue of valuation was of a financial, eco-
nomic, and technical nature).
203. Corfu Channel (U.K. v. Alb.), Judgment, 1949 I.C.J. Rep. 4, 13, 14, 21 (Dec. 15).
204. See Scovazzi, supra note 199, at 16
[T]here is no doubt that naval officers [such as those appointed in the Corfu Chan-
nel case] are better suited than anyone else to clarify whether the operation of
laying of mines in a given straight by a third subject can be seen and heard by the
look-out posts on the coast. Judges sitting in courts are not likely to engage them-
selves in activities of investigation on moonless nights, such as the test of visibility
carried out by the committee of experts in the Strait of Corfu.
205. NAOMI ORESKES & ERIK M. CONWAY, MERCHANTS OF DOUBT: HOW A HANDFUL
OF SCIENTISTS OBSCURED THE TRUTH ON ISSUES FROM TOBACCO SMOKE TO GLOBAL
WARMING 272 (2010) (“[W]e need to pay attention to who the experts actually are – by
asking questions about their credentials, their past and current research, the venues in which
they are subjecting their claims to scrutiny, and the sources of financial support they are
receiving.”).
ers with the opportunity to appoint additional experts or to require the

parties to disclose the limits inherent in their choice of experts.
To supply testimony that can be considered reliable, an expert witness
must have more than general knowledge of the topic on which they are
opining.206 They must have experience in the particular type of problem
at issue in a dispute,207 partly because so many tasks in the real world do
not “correspond to the specialty of any single profession.”208 Experience
in advising on corporate acquisitions of oil field properties, for example,
would not mean that a witness is qualified to testify on other aspects of the
oil and gas industry, such as the complex hedging of risk through futures
trading.209 Nor is an appraiser “at home in the whole field of appraisal[:]
he is a real-estate appraiser, a mine appraiser, a security analyst.”210
As for what counts as experience in a reliability analysis, what matters
is an expert’s previous experience with the chosen methodology outside
the dispute context. At the same, an expert need not have conducted or
supervised the specific studies on which her testimony relies.211 If the
question is whether certain human activities can cause damage to coral
reefs, for example, an expert is more likely to provide reliable answers to
that question if she has experience in evaluating anthropogenic effects on
coral reefs as part of an academic or other study, although she need not
have studied the particular coral reef at issue in the dispute.212 An expert
asked to opine on the considerations that go into construction of large
dams is more likely to provide reliable answers if she has real-world expe-
206. Prosecutor v. Peris̆ić, Case No. IT-04-81-T, Decision on Defense Motion to Ex-
clude Expert Reports of Robert Donia, ¶ 8 (Int’l Crim. Trib. For the Former Yugoslavia Oct.
27, 2008) (“[S]tatements or reports of an expert witness will only be treated as expert evi-
dence, insofar as they are based on the expert’s specialised knowledge, skills or training.
Statements that fall outside the area of expertise will be treated as personal opinions of the
witness and will be weighted accordingly.”); see also ORESKES & CONWAY, MERCHANTS OF
DOUBT, supra note 205, at 271 (“An all-purpose expert is an oxymoron.”).
207. See WHITE, supra note 9, at 148 (noting that, “[i]n the field of valuation of prop-
erty, the qualified expert is likely to be a specialist in a particular type of valuation
problem”).
208. See JAMES C. BONBRIGHT, 1 THE VALUATION OF PROPERTY 7 (1937) (considering
the task of asset valuation).
209. See Proton Energy Group SA v. Orlen Lietuva, (2013) EWHC (Comm) 2872, QB
(Eng.).
210. See BONBRIGHT, supra note 208, at 7.
211. The WTO Appellate Body has come to the opposite conclusion in at least one
case. See US — Continued Suspension, supra note 111, ¶ 199 (discounting an expert’s testi-
mony in part because the underlying scientific studies were not “carried out by him under his
supervision”).
212. See Phil. v. China, Case No. 2013-19, Award, ¶ 821 (Perm. Ct. Arb. 2016) (noting
that a tribunal-appointed expert asked to opine on the potential effects of China’s construc-
tion of man-made islands on surrounding coral reefs was a “coral reef ecologist with over ten
years’ research experience in Southeast Asia [and] the Pacific Islands . . . and ecological
work . . . focused on coral reef restoration and ecological functioning and the impact of
environmental and anthropogenic factors on coral reef benthic communities”).
rience in the construction of large dams, although she need not have par-
ticipated in the construction of the dam at issue in the dispute.213
Any assessment of an expert’s reliability must also include a thorough
examination of any previous participation as an expert in other legal dis-
putes.214 The fact that the expert’s previous participation in other disputes
was consistent with the evidence that she provides in the dispute at hand
and that other legal decision makers relied on an expert’s use of the same
or similar methodology can be an indicator of reliability, although no sub-
stitute for a decision maker’s own reliability analysis.
Confidentiality associated with some forms of international dispute
settlement, including commercial arbitration, may make it difficult to as-
sess reliability on the basis of an expert’s prior participation in interna-
tional disputes. But agreements and procedural rules that increasingly
emphasize the benefits of disclosing decisions, briefs, and witness reports
and statements may enable decision makers to more easily track and con-
sider the history of experts.215 And a decision maker can also of course
look to an expert’s participation in domestic court proceedings or domes-
tic regulatory rulemaking.
When the parties do not agree on the sort of expertise required, a
decision maker may decide to appoint (in the case of tribunal-appointed
experts) or weigh (in the case of party-appointed experts) a wider range of
experts who reflect those disparate views.216 But a decision maker should
remain aware that the parties’ views may reflect strategic choices intended
213. See Harza v. Iran, Interlocutory Award, 2 Iran-U.S. Cl. Trib. Rep. 68, 71 (1983)
(determining, in a case concerning the site-selection for and construction of major dams and
related irrigation networks “will require opinions to be sought at least from two experts, one
experienced in the field of geology and rock mechanics and the other in hydraulics, both of
whom are also experienced in the building of major dams” (emphasis added)).
214. See, e.g., Prosecutor v. Peris̆ić, Case No. IT-04-81-T, Decision on Defense Motion
to Exclude Expert Reports of Robert Donia, ¶ 12 (Int’l Crim. Trib. For the Former Yugosla-
via Oct. 27, 2008).
215. For example, mechanisms such as the United Nations Convention on Transparency
in Treaty-based Investor-State Arbitration, through application of the UNCITRAL Rules on
Transparency to a potentially large set of investment arbitrations, may increase the database
of expert reports and international decisions relying on expert evidence. See G.A. Res. 69/
116 (Dec. 10, 2014) (applying the UNICTRAL Rules on Transparency to investor-state dis-
putes where both the respondent and the claimant’s state are parties to the Convention, and
the treaty was concluded before April 1, 2014); G.A. Res. 68/109 (Dec. 16, 2013) (providing
that “expert reports and witness statements, exclusive of the exhibits thereto, shall be made
available to the public, upon request by any person to the arbitral tribunal”).
216. See Panel Report, EC — Asbestos, ¶ 5.2 WTO Doc. WT/DS135/R and Add. 1
(adopted 5 April 2001) (describing, in a dispute over whether a French measure banning the
marketing of asbestos products were inconsistent with WTO rules, the Canadian preference
for experts specializing the comparative toxicity of asbestos and non-asbestos fibers, and the
EC preference for cancer epidemiologists, including specialists in asbestos-caused cancers);
US — Continued Suspension, supra note 111, ¶ 7.91 (describing the EC’s view that the panel
should have appointed an expert in animal science, including veterinary practices on adminis-
tration of growth hormones).
to influence the outcome of a dispute,217 and may not encompass the full
range of expertise necessary to evaluate the questions presented.
3. What methodologies does the relevant technical field consider

reliable to answer the question presented?
Once a question has been formulated and a decision maker has as-
sured itself of an expert’s qualifications, the next step is to determine
whether the expert’s methodology is considered reliable in the relevant
technical field. Most technical fields rely on “peer review”—evaluation by
other specialists in the field—to determine whether work is potentially
valid.218 The fact that a methodology has been subject to and withstood
the scrutiny of peer review can signal that other experts in the relevant
field have analyzed the methodology, determined that the methodology is
capable of reliable application, and generally view the methodology as
credible.219 At base, peer review determines “[w]hat counts as knowl-
edge” in technical and specialized fields.220 The goal of peer review is to
assess the quality of analysis, including the design and execution of any
studies on which that analysis is based,221 and to identify gaps and limits in
knowledge and the need for further investigation.222
Because “peer review provides a benchmark of legitimacy,” it can pro-
vide a basis for assessing the reliability of expert evidence in international
disputes.223 As the need to rigorously examine the reliability of expert
217. See Gruszczynski, supra note 190, at 228 (“[A] preference for a specific type of
expertise may reflect deliberate and strategic choices of the parties aimed at influencing the
ultimate outcome of a dispute.”).
218. See Orellana, supra note 180, at 65 (“Questions of scientific truth are the function
of peer review, which serves as a symbol of professional accountability and ensures the dem-
ocratic control of science.”).
219. See Orellana, supra note 180, 57 (“[A]lthough the political nature of the Intergov-
ernmental Panel on Climate Change has been evident since its creation, the institutional
mechanisms designed for discussion between scientists and policy makers, coupled with ex-
tended peer review, have significantly contributed to the panel’s credibility.”).
220. See ORESKES & CONWAY, MERCHANTS OF DOUBT, supra note 205, at 269 (noting
that peer review determines what scientists consider to be scientific knowledge).
221. See FOSTER, supra note 3, at 10.
222. See Wirth, supra note 133, at 842
[T]he scientific peer review process operating in a regulatory context can reduce
disagreement, identify gaps and holes, and articulate the need for further investiga-
tion. . . . [P]eer review is responsive to a characterization of science as an ongoing
search for knowledge against a constantly shifting and evolving background that by
its very nature is always operating at new frontiers.
223. See Orellana, supra note 180, at 65 (“[T]ribunals facing the difficult task of recog-
nizing the scientific character of evidence will find that peer review provides a benchmark of
legitimacy that justifies qualified deference.”); Jacur, supra note 181, at 452 (“Peer review is,
in my view, a reliable instrument to ensure the credibility of experts.”); MARBOE, supra note
64, at 183 (explaining that, in investment arbitrations, “an arbitral tribunal needs basic
knowledge about the most important valuation principles and methods which are known and
applied in international valuation practice”); cf. Céline Lévesque, Science in the Hands of
International Investment Tribunals: A Case for “Scientific Due Process”, in 20 FINNISH Y.B.
evidence has increased with the complexity of international disputes over

the years, legal decision makers have begun to rely more heavily on the
relevant technical communities for guidance.224 This has mirrored a simi-
lar trend in the domestic law context.225 Parties to international agree-
ments now hardwire reliance on technical communities into the legal text
itself.226
But peer review is far from perfect,227 and a legal decision maker
should insist that each expert demonstrate that the testimony is supported
by peer-reviewed primary materials and that those materials as a whole
meet minimum acceptable standards for reliability in the field,228 which
will generally mean that the research was conducted in a disinterested
manner, shared with other researchers, and subjected to disinterested
evaluation by peers.229 Decision makers should analyze the reliability of
INT’L L. 259, 288 (2011) (“[A] tribunal might consider whether . . . regulatory science was
peer reviewed.”).
224. See e.g., Philip Morris Brands Sàrl v. Oriental Republic of Uruguay, ICSID Case
No. ARB/10/7, Award, ¶ 392 (2016) (noting, as detailed by Uruguay’s experts, that evidence
of consumers’ misconceptions of health effects of “light” and “low-tar” cigarettes available at
the time Uruguay adopted the challenged measures included “[n]umerous scientific stud-
ies . . . published . . . in leading international journals” and “cited by the U.S. Surgeon Gen-
eral and the U.S. National Cancer Institute” (citations omitted)); Whaling in the Antarctic
(Austl. v. Japan: N.Z. intervening), Judgment, 2014 I.C.J. Rep. 226 (Mar. 31) (separate opin-
ion of Bhandari, J.) (recognizing peer review as an indicator of “reliable science”); Methanex
Corp. v. United States, Final Award of the Tribunal on Jurisdiction and Merits, ¶ 101
(NAFTA Ch. 11 Arb. Trib. 2005), http://www.italaw.com/sites/default/files/case-documents/
ita0529.pdf (noting that an environmental impact report had been subject to “public hear-
ings, testimony, and peer review,” “emerging as a serious scientific work from such an open
and informed debate”).
225. See THOMAS O. MCGARITY & WENDY E. WAGNER, BENDING SCIENCE: HOW SPE-
CIAL INTERESTS CORRUPT PUBLIC HEALTH RESEARCH 1 (1st ed. 2008) (“[L]egal decision
makers have struggled to develop more rigorous tools for assessing the reliability of the sci-
entific information that informs health policies. The solution, most have agreed, is for deci-
sion makers to rely more heavily on the scientific community for oversight and assistance.”).
226. Foster, supra note 45, at 151 (reasoning that the “risk assessment” provision in the
SPS Agreement requires a WTO dispute settlement panel to “assess whether the science
relied on by a Member is legitimate science according to the standards of the relevant scien-
tific community and also scrutinize the reasoning in a risk assessment with reference to the
underlying science”).
227. See generally John P.A. Ioannidis, Why Most Published Research Findings Are
False, 2 PLOS MED. 696, 700–01 (2005); William A. Wilson, Scientific Regress, FIRST THINGS
(2016), https://www.firstthings.com/article/2016/05/scientific-regress (observing that “[a]t its
best, science is a human enterprise with a superhuman aim” but noting “example after exam-
ple of how the human element of this enterprise harms and damages its progress, through
incompetence, fraud, selfishness, prejudice, or the simple combination of an honest oversight
or slip with plain bad luck”).
228. See Ioannidis, supra note 227, at 700-01.
229. See MCGARITY & WAGNER, supra note 227, at 47 (setting out these “informal
norms of good scientific practice”); see also Markus Wagner, Law Talk v. Science Talk: The
Languages of Law and Science in WTO Proceedings, 35 FORDHAM INT’L L.J. 151, 157 (2011)
(“[S]cientific inquiry is disinterested in the policy outcome that follows any revelation,
through publication or otherwise, of the results of the scientific inquiry.”).
underlying evidence in view of the totality of the evidence presented, with-

out focusing on the outcomes of particular studies.230
How can non-experts discern when peer-reviewed studies are relia-
ble? A widely-cited analysis of peer-reviewed scientific research concluded
that research findings are more likely to be true when study sample sizes
are larger (e.g., the relatively large randomized controlled trials in cardiol-
ogy) and the demonstrated effects of studied phenomenon are larger (e.g.,
the relatively large effects of smoking on cancer or cardiovascular dis-
ease).231 Similarly, research findings in a particular area are more likely to
be true when confirmatory studies have been performed (e.g., meta-analy-
ses of previous studies) and adhere to standardized conduct and reporting
practices (e.g., those that decrease the discretion and flexibility of individ-
ual researchers in designing and reporting studies).232
Certain sorts of evidence demand heightened scrutiny, particularly
those where financial and other interests may unduly influence the re-
search process. For instance, when experts rely on “policy-relevant” re-
search, technical knowledge that may influence government policy or
litigation outcomes,233 they should demonstrate that the relevance of the
outcomes did not compromise the research process.
An advocate of a particular policy outcome can direct the outcome of
policy-relevant research by commissioning studies and then exerting sig-
nificant control so that the studies work backward from the advocate’s
desired outcome.234 Or the advocate may simply re-package existing stud-
ies in a light more favorable to the advocate, including by cherry-picking
230. See Ioannidis, supra note 227, at 700–01 (“[M]ost research questions are addressed
by many teams, and it is misleading to emphasize the statistically significant findings of any
single team. What matters is the totality of the evidence.”); Christie Aschwanden, Science
Isn’t Broken: It’s Just a Hell of a Lot Harder than We Give It Credit For, FIVETHIRTYEIGHT
(Aug. 19, 2015), https://fivethirtyeight.com/features/science-isnt-broken/ (“The important les-
son here is that a single analysis is not sufficient to find a definitive answer.”).
231. Ioannidis, supra note 227, at 697–98.
232. See id.; see also Aschwanden, supra note 230 (describing the “decisions scientists
make as they conduct a study” that can unduly influence research outcomes).
233. See MCGARITY & WAGNER, supra note 225, at 10–11 (2008); see also Bonneuil &
Levidow, supra note 49, at 77 (“[N]either regulatory science nor research science can claim to
stand free from historical, political and cultural situations.”); Ioannidis, supra note 227, at 698
(“Conflicts of interest and prejudice may increase bias. Conflicts of interest are very com-
mon in biomedical research, and typically they are inadequately and sparsely reported. . . .
The hotter a scientific field . . . the less likely the research finding are to be true.”).
234. See MCGARITY & WAGNER, supra note 225, at 8–9, 38–39 (describing this phe-
nomenon in the context of scientific research); Werner, supra note 33, at 48
The distrust in expert input can partly be explained by the fact that scientific ex-
perts sometimes ally with specific commercial firms, interest[ ] groups and political
parties. One of the most disturbing examples are the so-called ‘merchants of
doubt,’ a group of scientific experts with strong links to industry and conservative
politicians who proved willing to cast doubt on results of scientific research in the
areas of smoking tobacco, acid rain, the depletion of the ozone layer and global
warming.
research.235 Because policy-relevant research often addresses narrow

questions arising from policy debates, advocates—especially regulated in-
dustries—may be the only ones interested in pursuing or reviewing it, or
to devote sufficient resources to it.236 This may lead to asymmetrical rep-
resentation of views in those areas, in favor of the outside advocates’ posi-
tion, as well as little disinterested examination of the details of the
research.237 Outside advocates know that funding policy-relevant re-
search can then receive the imprimatur of legitimacy through peer review
because lower-tier publications are eager for submissions and may subject
research to only cursory review.238
If it appears that research supporting expert evidence is “policy-rele-
vant,” the next question is whether outside advocates—with an interest in
the outcome—funded or otherwise participated in the research or peer-
review process. Although the mere fact that advocates of a particular out-
come have funded policy-relevant research does not mean that it is unreli-
able,239 in many cases advocate funding does influence research
outcomes.240 For that reason, advocate funding, especially when coupled
with contractual control over research results or publication, or an advo-
cate’s significant collaboration in the research process, require a legal deci-
sion maker to look further into the provenance of empirical evidence. At
a minimum, generally accepted norms require disclosure of research spon-
sorships or other potential financial interests in the subject matter, and
confirmation that the named authors of a publication directed and con-
ducted the underlying research.241 A tribunal or court should allow the
235. See MCGARITY & WAGNER, supra note 225, at 39.

236. See id. at 11–13, 54–55.
237. See id. at 54–55 (“[S]tudies that sponsors control in obvious ways may attract little
attention from independent scientists because they are reluctant to spend their valuable time
refuting studies that most of their peers know to be unreliable.”).
238. See id. at 53 (noting that “many second- and third-tier [scientific] journals are hun-
gry for submissions and will publish articles that receive minimal peer scrutiny both before
and after publication.” (citation omitted)).
239. See id. at 65
Independent scientists can and regularly do conduct policy-relevant research in a

disinterested way, without input from affected parties or financial inducements
that cause them to tilt or skew the research toward a particular end. Furthermore,
scientists can and do carry out sponsored research in accordance with the norms
and procedures of science without being influenced by the sponsors’ economic
interests.
(citation omitted).
240. See id. at 95–96 (detailing the “funding effect” in biomedical research, which ac-
cording to a “comprehensive review article summarizing 1,140 biomedical research studies”
means that “ ‘industry-sponsored studies were significantly more likely to reach conclusions
that were favorable to the sponsor than were non[-]industry studies.’ ” (citation omitted)).
241. See id. at 76–77:
Most scientists would agree that when a sponsor contractually controls research or
otherwise acts as a significant collaborator, the norms of science require that any
parties to present evidence on the extent to which advocates have commis-

sioned research underlying expert evidence and the degree to which advo-
cates controlled the research outcomes.242 If a key publication of policy-
relevant research does not disclose this information, the party relying on
the publication should provide those disclosures and explain why any ad-
vocate’s financial interest or other involvement did not lead to unreliable
results.243
Lastly, although peer review can be a significant indicator of reliabil-
ity, not all reliable work is peer-reviewed or even published.244 For spe-
cialized fields with significant industrial or commercial applications, for
example, standards from the non-academic private sector may not be
peer-reviewed but will often play an important role in defining whether a
particular methodology is generally accepted.245
published version of the study disclose at least the fact of sponsorship . . . . The
most insidious technique for obscuring provenance is the practice commonly em-
ployed in the pharmaceutical industry of hiring ghost-writing companies that in
turn hire bright young science writers to turn data and analyses from company-
sponsored clinical trials into articles suitable for publication in medical journals
under the signatures of prominent researchers.
242. See id. at 240
The [U.S.] courts should allow the parties to conduct discovery and present evi-
dence on the extent to which some or all of the research underlying an expert’s
testimony was in fact commissioned by an entity with a direct or indirect interest in
the litigation and on the degree to which that entity exercised control over the
outcome of the research.
243. See id. at 239 (“[C]ourts and agencies should at the very least ensure that the prov-
enance of all research that they rely on to support regulatory and judicial determinations is,
as far as is possible, divulged and considered in the decision-making process.”); Biases and
errors may stem from sources other than advocate funding. Specialists in a particular field
may have biases due to their commitments to certain theories or to certain scientific organi-
zations. Legal decision makers should take into account whether any potential biases from
these other sources have undermined the reliability of the expert’s conclusions. Ioannidis,
supra note 227, at 698 (“Prejudice may not necessarily have financial roots. Scientists in a
given field may be prejudiced purely because of their belief in a scientific theory or commit-
ment to their own findings.”); Theofanis Christoforou, Settlement of Science-Based Trade
Disputes in the WTO: A Critical Review of the Developing Case Law in the Face of Scientific
Uncertainty, 8 N.Y.U. ENVTL. L.J. 622, 630 (2000) (“[S]cientists coming from [specific] orga-
nizations may be unfairly biased in favor of maintaining their organization’s standards and
recommendations”); see also Gruszczynski, supra note 32, at 227 (“[P]revious affiliation
should be regarded as an important element when assessing the existence of potential con-
flicts of interest.”).
244. See Matthew W. Swinehart, Remedying Daubert’s Inadequacy in Evaluating the
Admissibility of Scientific Models Used in Environmental-Tort Litigation, 86 TEX. L. REV.
1281, 1307 (2008) (cautioning against similar misuse of peer review as a factor in analyzing
expert reliability).
245. FOSTER, supra note 3, at 9 (“Standards emanating from the private sector may also
have an effect within international law.”).
4. Which expert’s methodology is most reliable in this context?

Once a tribunal has satisfied itself that the methodology is one that is
generally accepted and used in the field to address the sort of question at
issue, a reliability analysis should then ask whether the expert has chosen a
methodology that appropriately fits the particular circumstances of the
case and has explained why that methodology is preferred over other
available or proffered methodologies.246
Views of the Relevant Technical Field—To use expert evidence to
make decisions, decision-makers need not arrive at a unified theory by
synthesizing or reconciling competing expert evidence, but must assess the
relative reliability of each expert submission.247 This is critical given that
an evaluation of a methodology’s reliability will rarely if ever lend itself to
a binary analysis, resulting in a definitive determination that particular
methodologies are “reliable,” while others are “not reliable.” Instead, the
reliability of expert methodologies is a matter of degree, and the degree of
each methodology must be compared to that of others.248 To that end, a
decision maker should identify and examine the generally accepted norms
and standards of the expert’s field,249 including the relative reliabilities of
246. See Pauwelyn, supra note 77, at 186 (“[T]he pros and cons [of empirical economic
methodologies] should be made explicit, including potential statistical and identification
problems” and “[a]lternative methodologies should be discussed.”); RIPINSKY & WILLIAMS,
supra note 43, at 192 (“There is no single valuation method that fits all cases and scenarios.
This gives tribunals flexibility but poses the problem of selecting a method that is appropriate
and justified in a particular case.”)
247. See, e.g., Panel Report, US — COOL, ¶ 7.513 , WTO Doc WT/DS384, 386/R (Nov.
18, 2011) (noting, in evaluating competing economic studies, that it is not the task of a WTO
panel, “to establish a unified econometric report or to conduct our own econometric assess-
ment; instead we will assess the robustness of each study”); CMS Gas Transmission Company
v. Argentine Republic, ICSID Case No. ARB/01/8, Award, ¶¶ 411–417 (May 12, 2005) (ex-
plaining why the tribunal chose to value the business at issue using the DCF method rather
than other approaches such as the multiples, comparables, option-value, and asset-value
methods); Starrett Housing Corp. v. Iran, Final Award, 16 Iran-U.S. Cl. Trib. Rep. ¶¶ 282,
336 (1987) (noting approvingly that an expert employed “generally accepted accounting prin-
ciples” and “generally recognized valuation practices”).
248. See Pauwelyn, supra note 77, at 26 (“[T]he scope, relevance, assumptions and lim-
its of each study should be examined and, on that basis, each should be given appropriate
weight (or, as the case may be, no weight at all).”); cf. CHRISTOPH H. SCHREUER ET AL., THE
ICSID CONVENTION: A COMMENTARY 1012 (2d ed. 2009) (“The speculative character of
damages theories in the calculation of lost profits is a matter of degree.”).
249. See, e.g., EC—Asbestos supra note 216, ¶ 5.147 (determining that it was generally
accepted among scientists that there was “a direct and linear relationship between the rela-
tive risk of lung cancer and cumulative exposure to asbestos”); CMS Gas Transmission Com-
pany, ICSID Case No. ARB/01/8, Award, ¶ 416 (adopting the DCF valuation method in part
because it is “universally accepted”); Harza v. Iran, 2 Iran-U.S. Cl. Trib. Rep. at 76, ¶ 99
(1986) (declining to rely on an expert opinion that the claimant could have improved its
geological assessment in preparing for the construction of a dam because “the evidence indi-
cates, no clear technical or professional standards exist” as to the necessary elements of such
an assessment in that context). The idea that tribunals should consider the “most common
and accepted methods” in making damages assessments comes from the legal standard of
“fair market value.” See RIPINSKY & WILLIAMS, supra note 43, at 211 (“If one really wishes
to estimate the market value of an investment, then one should use most common and ac-
different methodologies as gleaned from peer-reviewed studies and col-

lected in national and international guidelines.250 A consensus reached
among decision-makers in international disputes on the use of a methodol-
ogy may also indicate that a methodology is the most reliable in certain
sorts of cases and fits within the framework of applicable legal rules,251
although it is not a substitute for full reliability evaluation.252
The use of a particular methodology can be adjudged “correct” only
after sufficiently widespread acquiescence in the field over time, a “con-
sensus” that is always subject to revision. Many specialized fields by their
very nature do not generally offer definitive answers but reflect some de-
gree of uncertainty, due to imperfect knowledge or ever-developing opin-
ions,253 and center instead on continual evaluation and testing of
hypotheses that are understood to be valid only until they are dis-
proved.254 The relevant field may remain in a state of disequilibrium on
the acceptance of a methodology, with specialists holding a range of
cepted methods to reach that value, and the DCF method is an appropriate method.”); Sim-
mons, supra note 64, at 234 (“Because tribunals are seeking to determine market value, their
decisions should be informed by the real-world practices of willing buyers and willing sell-
ers.”); id. at 235 (arguing that tribunals should “abandon the practice of dismissing the DCF
method simply because an entity being valued is not a ‘going concern’ ” in part because
“[f]inancial analysts employ the method even for enterprise that arbitral tribunals would not
deem ‘going concerns’ ”). But even absent the legal standard the general acceptance of par-
ticular valuation methodologies has also informed broader considerations of the reliability of
valuation evidence. See, e.g., Manuel A. Abdala, Key Damage Compensation Issues in Oil
and Gas International Arbitration Cases, 24 AM. U. INT’L L. REV. 539, 548–49 (2009) (noting
that DCF is “the most common methodology used in valuation analyses” in part because “it
is widely supported by the professional literature” and “most investors rely on a DCF analy-
sis to determine whether or not to undertake a particular project” (citations omitted)); Star-
rett Housing, 16 Iran-U.S. Cl. Trib. Rep. at 241 (Holtzmann, J., concurring opinion)
(reasoning that “experience in the financial community throughout the world attests to the
suitability of the DCF Method in valuing all kinds of businesses regardless of their purposes
or the length of time they are expected to operate”).
250. See RIPINSKY & WILLIAMS, supra note 43, at 192 (noting that, although “[t]here is
no comprehensive and authoritative manual of universal application that tribunals could use
to guide them through valuation. . . . there are a number of well-respected academic manuals
as well as national and international guidelines on valuation that can be referred to.” (cita-
tion omitted)).
251. Cf. CMS Gas Transmission Company, ICSID Case No. ARB/01/8, Award, ¶ 416
(adopting the DCF valuation method in part because it has been adopted by “numerous
tribunals”); see also RIPINSKY & WILLIAMS, supra note 43, at 192 (“[T]he growing body of
arbitral practice need[s] to be reviewed and analysed in order to understand the content and
applicability of different valuation methods for the purposes of damages awards in invest-
ment arbitrations.”).
252. See supra notes 138–39 regarding the discount rate discussion in Venezuela Hold-
ings, and accompanying text.
253. See Miles, supra note 33, at 157 (“[S]cience is not always able to provide exact,
conclusive answers, particularly where there is uncertainty, imperfect knowledge or conflict-
ing scientific opinion.”).
254. See id. (noting that “scientific knowledge” is “an organic process of evolving un-
derstandings and continuous interactive exchanges of hypothesis and refutation”); Douglas
Crawford-Brown et al., Environmental Risk, Precaution, and Scientific Rationality in the Con-
text of WTO/NAFTA Trade Rules, 24 RISK ANALYSIS 461, 462 (2004) (“An important aspect
views.255 Consensus may simply remain elusive on a time-scale useful to

legal decision makers, who must make decisions today, especially when it
comes to cutting-edge scientific research.256
For that reason, expert evidence need not reflect “mainstream” or
consensus opinion to be reliable. The methodology—not the conclusion—
should be generally accepted, even if an expert’s conclusions based on that
methodology diverge from the opinions of most other experts in the
field.257 The WTO Appellate Body has reached a similar conclusion, in a
slightly different but analogous context: how to determine when a govern-
ment has “based” a health measure on a “risk assessment,” consistent with
Article 5.1 of the SPS Agreement. The Appellate Body has consistently
explained that a government may base a risk assessment not only on main-
stream science but also on scientific opinions outside the mainstream.258
At the same time, it has required minority opinions to “have the necessary
scientific and methodological rigour to be considered reputable sci-
of science is that it is by nature progressive, with methods designed to approach a better

understanding over time.”).
255. See Naoto Jinji, An Economic Theory of the SPS Agreement, in SCIENCE AND
TECHNOLOGY IN INTERNATIONAL ECONOMIC LAW: BALANCING COMPETING INTERESTS 53,
54 (B. Mercurio & Kuei-Jung Ni eds., 2014) (“[T]he scientific community may fail to reach a
consensus [in risk analysis regarding health and safety], as a number of opposing but equally
plausible views may exist.”); MCGARITY & WAGNER, supra note 225, at 56 (“[I]n some areas
of public health and environmental research, scientists are in no position to validate the accu-
racy of a particular study’s methodological approach at the time its findings are announced
and pressed on agencies or courts.”).
256. See Orellana, supra note 180, at 55
While scientific method and peer review lead towards consensus as science’s oper-
ational criteria, consensus, instead, becomes more elusive the farther out from core
knowledge and the farther into the frontier areas of research and experimentation
one travels. Such frontier areas are characterized by persistent uncertainty and
diverging interpretations, where emerging consensus or minority views may de-
pend on socio-cultural factors and on informal negotiations among scientists.
(citations omitted).
257. See Pauwelyn, supra note 48, at 252–54 (reasoning, in the context of WTO disputes
over health policy, that there is “no need to find that at least a majority of the scientific
community is in favour of a proposed health measure,” even though “the minority opinion
must nonetheless come from ‘qualified and respected sources’ ”); Caroline E. Foster, The
“Real Dispute” in the Southern Bluefin Tuna Case: A Scientific Dispute?, 16 INT’L J. MARINE
& COASTAL L. 571, 589 (2001) (“Developments in multilateral international trade law also
recognise that the key features of what may constitute genuine ‘science’ do not necessarily
include a requirement that a scientist belong to a ‘mainstream’ scientific community, depend-
ing on the circumstances. There is an increased emphasis on scientific principles and
methodology.”).
258. See, e.g., US — Continued Suspension, supra note 111, ¶ 591; see also Lukasz
Gruszczynski, Science and the Settlement of Trade Disputes in the World Trade Organization,
in SCIENCE AND TECHNOLOGY IN INTERNATIONAL ECONOMIC LAW 11, 13 (Bryan Mercurio
& Kuei-Jung Ni eds., 2014) (“The SPS case law is consistent in holding that risk assessment
may be based not only on mainstream science, but also on the opinions of scientists taking a
divergent view.”).
ence.”259 In other words, as one commentator has noted, consideration of

minority opinions should involve an “evaluation of a specific claim against
more general theories or established scientific views, the logical coherence
of a claim, its methodological soundness, or its ability to explain a particu-
lar phenomenon.”260
Failure to evaluate in a transparent manner the relative reliability of
competing expert evidence can leave a decision vulnerable to criticism. In
Philip Morris v. Uruguay, tobacco companies challenged Uruguay’s ciga-
rette marketing restrictions, including “single presentation” requirements
that prevented the companies from selling one brand (e.g., Marlboro)
under more than one variant (e.g., “Light” and “Blue”).261 Both sides
introduced expert evidence on the tobacco companies’ claim that the re-
quirement was overbroad because, while descriptive brand variants such
as “Low Tar,” “Light,” or “Mild” might lead consumers to believe incor-
rectly that some cigarettes are less harmful than others, other brand vari-
ants—such as “Red,” “Gold,” and “Blue”—were not similarly
misleading.262
Two arbitrators rejected the companies’ challenge to the single presen-
tation requirement, partly relying on Uruguay’s expert evidence and an
amicus submission from the World Health Organization (WHO) that as-
serted that brand variants, including different colors, could mislead con-
sumers.263 The third arbitrator, concurring in part and dissenting in part,
found the competing expert evidence, from the tobacco companies, more
compelling in key respects.264 None of the arbitrators, however, explained
why it viewed one set of expert evidence as more reliable than the other.
In concluding that the measure was “not disproportionate” to the public
health concern,265 the majority offered in passing that the “rationale” for
the requirements was “supported by the evidence” “as held by the
WHO”266 and that Uruguay’s experts had “indicated” that the tobacco
companies’ promotion of “light” cigarettes as a way to reduce smoking
risk “misrepresented what would actually happen.”267
The third arbitrator concurred in the majority’s judgment that Uru-
guay’s prohibition of descriptive and color brand variants was not dispro-
portionate, finding Uruguay’s evidence on color variants “sufficient,”
259. See US — Continued Suspension, supra note 111, ¶ 591.

260. See Gruszczynski, supra note 258, at 13.
261. See Philip Morris Brands Sàrl v. Oriental Republic of Uruguay, ICSID Case No.
ARB/10/7, Award, ¶¶ 357–367 (2016).
262. Id. ¶¶ 391–392, 406; see also id. ¶ 160 (Born, J., concurring in part and dissenting in
part).
263. See id. ¶ 407 (determining that, “as held by the WHO, ‘the rationale for [Uru-
guay’s] action [was] supported by the evidence’ ” (quoting WHO submission)).
264. See id. ¶¶ 159–164 (Born, J., concurring in part and dissenting in part).
265. Id. ¶ 409.
266. Id. ¶ 407.
267. Id. ¶ 403.
albeit “very limited,” unimpressive,” and “tenuous.”268 The concurrence

did not discuss the content of any of the expert reports or provide substan-
tive reasons for these characterizations except to say that they were based
on “the reasons detailed in the Claimants’ expert evidence.”269
The third arbitrator dissented with respect to all other brand variants,
such as words or numbers in addition to the brand name, seasonal or geo-
graphic variations, or different languages.270 But again the dissenting por-
tions of the opinion disclose nothing about the substance of the expert
reports, concluding summarily that “the fact that some uses of colors in
some brands of tobacco products may be regarded as misleading in some
circumstances does not suggest, even indirectly, that all other variations
. . . are also misleading.”271 This analysis failed to address the WHO’s
assertions that any “brand extension” or “brand variant” can create mis-
leading consumer perceptions, based on peer-reviewed studies purport-
edly showing that “people try to find attributes among brand variants” and
that any feature of pack design can create misconceptions.272 From the
face of the arbitrators’ opinions, then, the parties and the public cannot
determine whether the arbitrators sufficiently evaluated the expert evi-
dence or the extent to which the evidence factored into their conclusions.
Tribunals in investments disputes have also often failed to evaluate the
relative reliability of competing valuation methodologies, particularly
when the investment has little or no track record and is thus not consid-
ered a “going concern.”273 The tribunal in BG Group v. Argentina, for
example, dismissed use of the DCF method as leading “to a result which is
uncertain and speculative,” without addressing the details of the expert’s
DCF methodology or considering whether specialists in the field would
have relied on that methodology in valuing the investment.274 Rather, the
tribunal rejected the reliability of the DCF methodology based on two
available transactions that the tribunal considered comparable to the in-
vestment.275 Tribunals retain broad discretion in determining the standard
268. Id. ¶ 161.

269. Id. ¶¶ 160–161.
270. Id. ¶¶ 162-163.
271. Id. ¶ 163.
272. See Philip Morris Brands Sàrl v. Oriental Republic of Uruguay, ICSID Case No.
ARB/10/7, Written Submission by the World Health Organization and the WHO Framework
Convention on Tobacco Control Secretariat, ¶¶ 77–81 (Jan. 28, 2015).
273. See Simmons, supra note 64, at 226–32 (outlining four decisions between 2007 and
2009 that rejected the DCF method on the ground that it was too speculative given available
evidence).
274. BG Grp. Plc. v. Republic of Argentina, Final Award, ¶ 439 (UNCITRAL Arb.
Proc. 2008), http://www.italaw.com/sites/default/files/case-documents/ita0081.pdf.
275. See id. ¶¶ 440–444; see also Siag v. Arab Republic of Egypt, ICSID Case No.
ARB/05/15, Award, ¶¶ 566–570 (May 11, 2009) (rejecting the DCF method in favor of the
comparable transaction method without comparing relative reliability of the methods).
for evidence underlying damages calculations,276 but a more robust and

defensible analysis would have considered the reliability of the DCF
method side-by-side with the reliability of the comparable transaction
method, given the use of those methods by financial analysts in the real
world and the available evidence underlying each method.277
Determining the relative reliability of each methodology does not pre-
clude the possibility that multiple methodologies may be reliably useful to
decision making, particularly where multiple methodologies can serve to
confirm the reliability of another. The investment tribunal in CME v.
Czech Republic, for instance, used an adjusted DCF calculation, explained
in considerable detail in the award, “as a confirmation of the Tribunal’s
findings” based on an offer made by a potential buyer of the investment a
six months before the treaty violation.278
Learning to Live with Uncertainty—Examining the relative degrees of
reliability of competing methodologies may lead to an understanding that
each has its own degree of uncertainty. But that examination does not
lead inextricably to the conclusion that any amount of approximation ren-
ders a methodology speculative and unreliable. Many technical fields op-
erate in full knowledge that conclusions are—necessarily—
approximations and not unassailable facts.279 As many courts and tribu-
nals in international disputes have recognized, uncertainty is an inextrica-
ble part of many technical fields, but the presence—and
acknowledgment—of uncertainty in an expert’s conclusion does not mean
that the methodology used to reach that conclusion is unreliable.280 At
the same time, legal decision-makers are routinely asked to make deci-
276. See International Law Commission, Draft Articles on Responsibility of States for
Internationally Wrongful Acts art. 36(2), 2001. Investment treaties generally provide little
guidance on the issue. See Simmons, supra note 64, at 230 n.203.
277. If these tribunals had asked, they may have determined that their assumptions
about the limits of DCF outside of the “going concern” context were misplaced. See Sim-
mons, supra note 64, at 235 (“Financial analysts employ the method even for enterprises that
arbitral tribunals would not deem ‘going concerns.’ ”).
278. See CME Czech Republic B.V. v. Czech Republic, Final Award, ¶ 604 (UNCI-
TRAL Arb. Proc. 2003), http://www.italaw.com/sites/default/files/case-documents/
ita0180.pdf.
279. See CATHERINE BUTTON, THE POWER TO PROTECT: TRADE, HEALTH AND UNCER-
TAINTY IN THE WTO 131 (2004) (“[T]he appropriate handling of uncertainties is part of the
scientific process of risk assessment.”); Wirth, supra note 133, at 837 (“Because science is
incomplete, the scientific data set underlying any regulation is necessarily incomplete. That,
however, does not diminish the scientific nature of the inquiry.”).
280. See, e.g., Himpurna Cal. Energy Ltd. v. PT (Persero) Persusahaan Listruik Negara
(Indon.), Final Award, ¶ 376 (May 4, 1999), 25 Y.B. COMM. ARB. 13 (2000) (“There is no
reason to apologise for the fact that [the DCF] approach involves approximations; they are
inherent and inevitable. Nor can it be criticized as unrealistic or unbusinesslike; it is precisely
how business executives must, and do, proceed when they evaluate a going concern. The fact
that they use ranges and estimates does not imply abandonment of the discipline of economic
analysis; nor, when adopted by arbitrators, does this method imply abandonment of the disci-
pline of assessing the evidence before them.”); MARJORIE M. WHITEMAN, 3 DAMAGES IN
INTERNATIONAL LAW 1694, 1699 (1943) (quoting Delagoa Bay and East African Railway
Company, Award of 30 May 1900, regarding reliance on calculations of future income while
sions based on probabilities.281 The focus should be on minimizing any

potential error by examining the relative appropriateness of the methodol-
ogy as compared to alternatives, the validity of the assumptions, and the
reliability of the information underlying the expert’s testimony.
Uncertainty can arise because the technical discipline is not suffi-
ciently evolved to explain or describe the subject at issue, or uncertainty
might arise simply because the necessary empirical evidence is unavaila-
ble.282 And in every legal dispute, an expert’s application of a methodol-
ogy to a particular set of circumstances requires “interpretation of data,”
“conjectures on projected consequences,” and individualized views on the
“appropriate responses” to those data and consequences.283
What if an applicable legal or evidentiary rule demands a particular
degree of certainty before an interest becomes legally protected or evi-
dence is considered admissible?284 Sergey Ripinsky and Kevin Williams
have posed this question in response to criticisms of the use of the DCF
method investment arbitrations.285 They confront an apparent “contradic-
tion” between “the two key principles in the law of damages: (a) the prin-
ciple of prohibiting the award of speculative damages, and (b) the
principle requiring the award of the ‘fair market value’ of the invest-
ment.”286 As with other subjects of expert testimony, “[s]peculation and
acknowledging that “such a computation made in advance on the basis of purely theoretical
data cannot hope to be absolutely accurate but only comparatively likely”).
281. Alvarez, supra note 64, at 87–88 (citing investment disputes and international
criminal cases in which “the tribunals were asked to make determinations involving
probabilities and not verifiable certainties” and noting that “[d]eciding on the basis of uncer-
tainty is what international and domestic courts do every day”).
282. See THE CONSERVATION FOUNDATION, RISK ASSESSMENT AND RISK CONTROL 5
(1985)
Virtually all elements of risk assessment are clouded with uncertainty, basically of
two kinds. First, the various scientific disciplines involved in assessing risk are not
sufficiently developed either to explain the mechanisms by which particular causes
produce particular effects or to provide good quantitative estimates of cause-and-
effect relationships. Second, the data needed to analyze particular risks are usually
not available.
See also Crawford-Brown, supra note 254, at 468 (“[U]ncertainty is an intrinsic and essential
characteristic of science that must be passed on to the policy and legal arenas. Failure to do
so, by insisting on a single estimate of risk . . . does not fully and truthfully describe the state
of science at any moment.”).
283. See Miles, supra note 33, at 157.
284. See, e.g., Compania de Aguas del Aconquija, S.A. v. Argentine Republic, ICSID
Case No. ARB/97/3, Award, ¶ 8.3.4 (Aug. 20, 2007) (requiring future losses to be proved with
“some level of certainty”); Autopista Concesionada de Venezuela, C.A. v. Bolivarian Repub-
lic Of Venezuela, ICSID Case No. ARB/00/5, Award, ¶ 351 (Sept. 23, 2003), (requiring future
losses to prove “sufficient (degree of) certainty”); see also International Law Commission,
supra note 276 (setting out a requirement that anticipated income streams to be proved with
“sufficient certainty” to become a legally protected interest).
285. RIPINSKY & WILLIAMS, supra note 43, at 211.
286. See id.
uncertainty” is “inherent in any DCF analysis” but “can be dealt with

through evaluating the constituent elements of the expert’s testimony, “by
taking conservative estimates of cash flow projects and application of a
higher discount rate.”287 Ripinsky and Williams ultimately conclude that
a legal rule that demands “market” value necessarily contemplates a de-
gree of uncertainty consistent with accepted practices in the relevant field,
noting that, “[i]f one really wishes to estimate the market value of an in-
vestment, then one should use most common and accepted methods to
reach that value, and the DCF method is an appropriate method.”288
Even if a decision maker determines that the applicable legal standard is
fixed and cannot be reconciled with an expert’s methodology and its appli-
cation to a dispute, it should still ask and answer the question of whether
the relevant technical field would consider the expert evidence as reliable
for the decision at hand.
Lastly, consensus among experts in the field or in decisional law on a
methodology can evolve over time, and legal decision makers should not
hesitate to take a fresh look at matters that were once considered settled,
where the evidence suggesting a shifting expert consensus.289 In most
technical fields, a dedication to empiricism means that what is considered
generally accepted can and will change over time, and differences in views
among specialists is likely to continue in varying degrees. When a particu-
lar hypothesis or assertion becomes generally accepted in a field notwith-
standing those differences depends on the norms of the field, but in
general occurs well before the hypothesis or assertion enters the unassaila-
ble “core” of knowledge in a field.290 One example is the choice between
simple interest and compound interest in calculating monetary damages.
Although it was perhaps once considered “settled” that compound interest
was not allowable,291 many more recent investment tribunals have consid-
ered it well-established that compound interest should be applied in calcu-
lating damages.292
287. See id.

288. See id.
289. See Foster, supra note 257, at 590 (“[T]he understanding that scientific knowledge
is not static . . . is vital for any international tribunal dealing with a scientific case.”).
290. See id. at 10–11 (“What we know was an ‘invulnerable core of scientific knowledge
ultimately consists of scientific claims that no scientist any longer challenges. . . . However,
differences of view among scientists as to the validity of hypotheses and assertions can be
expected to remain indefinitely in varying degrees.”).
291. Compare WHITEMAN, supra note 280, at 1997 (noting an apparent mid-century
consensus); see also RIPINSKY & WILLIAMS, supra note 43, at 382. But cf. Natasha Affolder,
Awarding Compound Interest in International Arbitration, 12 AM. REV. INT’L ARB. 45, 71–73
(2001) (“The authorities cited by Ms. Whiteman . . . fail to support the existence of a general
principle of international law against the awarding of compound interest. At most, it can be
said that the question of whether compound interest can be awarded is an unsettled question
before international tribunals.”).
292. See, e.g., Siag v. Arab Republic of Egypt, ICSID Case No. ARB/05/15, Award, ¶
595 (May 11, 2009) (noting claimants’ submission that, “since 2000, no [fewer] than 15 out of
16 tribunals have awarded compound interest on damages in investment disputes”); see also
Relying on Available Data—The method that is most reliable in a par-

ticular case will depend in large part on the available data.293 In determin-
ing the fair market value of an investment, for example, experts have
employed a variety of methodologies, including accounting for the “book
value” of an investment, calculating the amount invested, assessing trans-
actions comparable to the investment, calculating future cash flows of the
investment according to the DCF method,294 or surveying a business’s
sales history of expropriated inventory.295 The reliability of valuation
methods that rely on prices actually paid for comparable property, for in-
stance, depends significantly on the availability of recent and representa-
tive comparable sales.296 For that reason, tribunals have relied more
readily on comparable sales when valuing real property, where representa-
tive comparable sales are more readily available, but have shied away
from relying on such data when valuing entire businesses.297
5. Has the expert presented the evidence in a way that is useful in

conveying information to a non-expert and clearly identifies
key points of disagreement with other experts?
The predictability and legitimacy of legal decisions depends in large
part on the transparency of the decision making process.298 But when it
comes to expert evidence, decision makers do not bear the burden of
transparency alone. The experts—who are involved in disputes to assist
the decision maker in making a more well-informed and accurate deci-
sion—also have a role in ensuring that they present evidence in a transpar-
ent manner, consistent with the principles and practices of their field.299
A decision-maker should therefore demand that experts convey their
RIPINSKY & WILLIAMS, supra note 43, at 387 (describing compound interest as “com[ing] to
be treated as the default solution”).
293. See MARBOE, supra note 64, at 205 (reasoning that the most reliable valuation
method “depends . . . on the question of whether a tribunal finds that the main parameters of
the method are appropriate and acceptable”).
294. See Simmons, supra note 64, at 223–24.
295. See MARBOE, supra note 64, at 195.
296. See Kardassopoulous v. Republic of Georgia, ICSID Case Nos. ARB/05/18, ARB/
07/15, Award, ¶ 598 (2010) (“It is not common in investment treaty arbitrations that a Tribu-
nal has available to it three arm’s-length, contemporaneous transactions (or potential trans-
actions) to assist in valuing an investment, much less three that converge in a narrow range of
value.”); see also Phelps Dodge v. Iran, 10 Iran-U.S. Cl. Trib. Rep. 121, ¶¶ 29–30 (1986)
(rejecting valuation by multiples because the consultant’s study relied on U.S. enterprises
that were not sufficiently comparable to a start-up business in Iran).
297. See MARBOE, supra note 64, at 201–02 (noting that “the Iran-US Claims Tribunal
frequently employed special experts in order to provide information about prices actually
paid in the real property market” but that “the comparable sales approach has not been
applied with respect to entire investment projects or undertakings in internationals invest-
ment disputes”).
298. See Walker, supra note 61, at 165.
299. See id. at 165–66 (noting that with respect to transparency “law and science share
the same ideal”).
opinions in a way that is readily accessible and understandable to non-

experts such as the decision-makers and future readers of the decision
maker’s conclusions.300
As an initial matter, a decision maker should ensure that each expert
has clearly disclosed all key assumptions and facts underlying the evi-
dence. This is the first step in understanding how an expert has applied
the chosen methodologies to the specifics of the case at hand. It also helps
to counteract an inherent asymmetry between experts, who know and un-
derstand a methodology’s inputs, and non-experts, who are unlikely to
know and understand those inputs without disclosure and explanation.301
Many evidentiary guidelines and rules now provide that experts should
make such disclosures. The IBA rules, for example, provide that an expert
should disclose a statement of the facts on which he or she is basing his or
her opinions and a description of the method, evidence, and information
used in arriving at the expert’s conclusion.302 Mark Kantor has proposed
that investment arbitrators impose on party-appointed experts a duty to
provide “full information” to the adjudicators, whether that information
supports or contradicts the expert’s professional analyses and conclusions,
and a duty to use assess, to the extent an expert has the background to do
so, the reasonableness of assumptions provided by counsel or a party.303
And Joost Pauwelyn has argued that parties relying on empirical economic
data should disclose all assumptions and that adjudicators should transpar-
ently evaluate those assumptions in their decisions.304
In this regard, one indicator that an expert has correctly applied a
methodology is disclosure of a transparent and flexible roadmap of under-
lying assumptions so that a tribunal may choose which of an expert’s as-
sumptions to accept and which to reject.305 Take, for example, the expert
300. See Foster, supra note 3, at xiv (“The science must be put forward in a form that is
readily digestible by a court or tribunal composed of individuals whose qualifications and
experience lie in the field of law rather than science.”).
301. Cf. Lorna Schrefler, Reflections on the Different Roles of Expertise in Regulatory
Policy Making, in THE ROLE OF “EXPERTS” IN INTERNATIONAL AND EUROPEAN DECISION-
MAKING PROCESSES: ADVISORS, DECISION MAKERS OR IRRELEVANT ACTORS? 63, 69–70 (M.
Ambrus et al. eds., 2014) (“Another example could be when data or assumptions in an eco-
nomic analysis are carefully selected to present a certain picture of reality, often exploiting
the information asymmetries between agents (in this case, the experts within the [regulatory
or administrative] agency) and principals.”).
302. IBA RULES 5.3(d)-(e); see also Wisner et al., supra note 94, at 243–44.
303. Mark Kantor, A Code of Conduct for Party-Appointed Experts in International
Arbitration – Can One Be Found?, 26 ARB. INT’L 323, 375 (2010).
304. Pauwelyn, supra note 77, at 187 (“Assumptions [underlying economic empirical
methodologies] must be disclosed and discussed. . . .”).
305. See Starrett Housing Corp. v. Iran, Final Award, 16 Iran-U.S. Cl. Trib. Rep.
112, ¶ 39 (1987) (“[T]he Expert referred many of his assumptions and decisions with regard
to his valuation to the Tribunal . . . [in order to] ma[k]e it easier for the Tribunal to under-
stand and possibly to adjust the final result according to the Tribunal’s own judgement.”); id.
¶ 269 (“[W]here [the expert] drew inferences or made subjective judgments, he pointed them
out and explained his reasons.”); see also Corfu Channel, (U.K. v. Alb.), Order, 1948 I.C.J.
Rep. 124, 126–27 (Dec. 17) (instructing a panel of three experts to “give the reasons for these
report in the Manica Plateau Arbitration. The report first laid out—in
plain language—the technical terms at the core of the parties’ dispute, and
how they were understood in geographical science.306 The expert then set
out to apply these “geographical principles” to the facts of the dispute and
the arguments of the parties, noting where the principles pointed one way
or the other.307 Only then did the expert turn to answering the specific
questions put to him by the arbitrator, laying out each incremental conclu-
sion by reference to the previously defined principles and analysis.308 By
setting out the relevant principles, assumptions, and facts, and connecting
them to his conclusions, the expert permitted the arbitrator to assess each
distinct line of analysis and modify as necessary.309
Outside the dispute context, experts commonly account for uncer-
tainty in their conclusions through statistical methods such as sensitivity
analyses, which describe how sensitive an expert’s conclusions are to
changes in underlying data, methodology, or assumptions. In disputes,
they also can assist decision makers in evaluating a range of inputs and
outcomes.310 This is especially true where small changes in assumptions
and other inputs can significantly alter an expert’s conclusions.311 A relia-
ble expert witness will disclose those assumptions and inputs, explain how
they drive the analysis, and identify material differences between that
analysis and the analyses of the other party’s experts. In Guarachi
America, Inc. & Rurelec PLC v. Bolivia, for example, the claimants’ valu-
ation expert provided an assessment that “95% of the difference” between
the parties’ valuations was due to three issues: discount rate, revenue pro-
findings in order to make their true significance apparent to the Court. . . . [and] mention any
doubts or differences of opinion amongst them”).
306. Manica Plateau Arbitration (Gr. Brit. v. Portugal), 28 R.I.A.A. 283, 302 (1897).
307. Id. at 302–03.
308. Id. at 303–07.
309. Id. at 308-09. The arbitrator reasoned that, although the expert’s conclusions were
“technically accurate,” the result was so “irregular” and full of “numerous inflections,” that it
could “easily give rise . . . to doubts and differences of opinion which should be carefully
avoided.” Id. at 308. As a result, the arbitrator asked the expert to modify one segment of
the proposed demarcation to “substitute[e] some nearly straight and better-defined lines for
the natural inflections . . . so that the extent of ground which each party gets by the substitu-
tion of straight lines for the rigorous demarcation of the edge remains almost equivalent.”
Id.
310. See Starrett Housing, 16 Iran-U.S. Cl. Rep. ¶¶ 39–40 (1987) (noting that tribunal-
appointed valuation expert “provided a sensitivity analysis for four essential areas to show
what effect their alteration would have on the valuation”); Pauwelyn, supra note 77, at 187
(suggesting that adjudicators require “thorough robustness analysis” of economic data used
in international disputes).
311. See Land Reclamation by Singapore in and Around the Straits of Johor (Malay. v.
Sing.), Verbatim Records, 37–38 (Sept. 25, 2003 a.m.), https://www.itlos.org/fileadmin/itlos/
documents/cases/case_no_12/PV.03.01.25.09.03.a.m.E.pdf (eliciting expert testimony that the
sediment transport rates at the core of the land reclamation dispute could change signifi-
cantly in response to only slight changes in velocity).
jections, and capital expenditures.312 The tribunal noted on several occa-

sions that each expert had disclosed critical assumptions, enabling the
other party’s experts—and the tribunal itself—to confirm or challenge
them.313 By contrast, an expert that requires a decision maker to accept
all underlying assumptions in order to rely on the testimony is unlikely to
provide valuable assistance and rightfully risks flat rejection.314
6. Was the methodology applied within practical and theoretical

boundaries?
It is of course not enough to determine that a method is reliable in the

abstract or that a witness is a legitimate expert in the methodology. A
tribunal must also ensure that the expert has reliably applied the method-
ology to the particular circumstances of the dispute at hand.315 This gener-
ally requires identifying the key premises and limitations of an expert’s
methodology and determining whether the assumptions and inputs that
have gone into the expert’s analysis are sound.
With respect to the key premises and limitations of an expert’s meth-
odology, a decision maker should confirm that the expert has applied the
methodology consistently, mindful of the methodology’s key premises and
limitations identified earlier in the reliability analysis. Reliable expert evi-
dence will not cherry-pick elements of a methodology that favor the ex-
pert’s conclusions but will instead faithfully apply the methodology within
its theoretical and practical limitations. If an expert has applied a peer-
reviewed methodology, citing certain publications or other sources, what
methodological limits do those sources recognize? What limitations do
other peer-reviewed sources recognize?
With respect to the assumptions and inputs underlying expert opinion,
a decision maker should ask whether they would be considered sufficiently
reliable in the relevant technical field316 and whether the expert has at-
tempted to verify all information and assumptions provided by the parties
312. Guarachi America, Inc. v. Plurinational State of Bolivia, Award, PCA Case No.
2011-17, ¶ 454 (Jan. 31, 2013).
313. See, e.g., id. ¶¶ 505–510 (noting differences in the experts’ assumptions regarding
future capacity of the investment, including assumptions related to demand and competition
in the market).
314. See Wisner et al., supra note 94, at 247–48 (citing an example of expert testimony
that permitted the decision maker to accept or reject each key assumption used in the ex-
pert’s valuation).
315. See RIPINSKY & WILLIAMS, supra note 43, at 194 (“[E]very valuation requires care-
ful analysis specific to the circumstances of the case.”).
316. See MARBOE, supra note 64, at 219 (emphasizing that the underlying “forecast of
future cash flows” in a DCF analysis must be “checked for plausibility in order to determine
its reasonableness and lack of contradiction”); Pauwelyn, supra note 77, at 187 (“Assump-
tions must . . . be consistent with the particular product, industry or market under examina-
tion.”); cf. Orellana, supra note 180, at 54 (noting that international adjudicators can evaluate
the reliability of scientific conclusions without deciding questions of “scientific validities” by
“resort[ing] to established techniques in fact-finding” and examining underlying evidence).
or the tribunal or court.317 If they are not, testimony that relies on such
data is not entitled to much, if any, weight.318
The tribunal’s analysis in Guarachi America, Inc. & Rurelec PLC v.
Bolivia provides a particularly useful illustration of this basic approach.319
The tribunal in that dispute expressly divided its analysis of the parties’
expert damages evidence into three primary areas of disagreement, identi-
fied the assumptions and inputs driving that disagreement, and compared
each assumption and input against the available documentary evidence.320
The result was a thoroughly reasoned valuation that sought to adopt the
most reliable assumptions and inputs from both parties’ experts and other
evidence in the record.321 The tribunal also looked to the primary sources
on which the experts relied to ensure that the methodology had been ap-
plied within its generally accepted boundaries. For instance, when Bo-
livia’s expert sought to apply a methodology that would have increased the
country risk associated with the investment, thereby increasing the dis-
count rate and lowering the valuation, the tribunal went back to the pri-
mary sources on which the expert had relied and discovered that those
sources would apply the methodology only in short-term scenarios (e.g.,
five to 10 years), not in the longer-term time period (28 years) at issue in
Guarachi America.322
When using the DCF method the assumptions and inputs critical to
the reliability of the valuation methodology can include past performance,
internal business plans and contract rights, and the regulatory framework
317. See Starrett Housing Corp. v. Iran, Award, 16 Iran-U.S. Cl. Rep. 112, ¶ 268 (1987)
(reasoning that it gave “substantial weight” to an expert report in part because of the “thor-
oughness of the process by which the Expert sought to verify all information presented to
him by the Parties”).
318. See, e.g., Arbitration Under Article 181 of the Treaty of Neuilly (“Certain Forests
in Rhodopia”), Decision on the Principal Question (Mar. 29, 1933), 28 AM. J. INT’L L. 760,
805 (1934). In Certain Forests in Rhodopia, Greece alleged that Bulgaria had dispossessed
Greek nationals of their rights in forest land during World War I. Id. at 760. The arbitrator
declined Bulgaria’s request for an expert assessment of the value of timber, reasoning that
the only evidence on which an expert could have relied – an in-person survey of the timber –
would have resulted in “speculative” testimony, in part because an expert could not have
taken into account any logging activities that had occurred in the 15 years since the alleged
dispossession. See id. at 805
The Arbitrator does not believe that it would be practical at the present time to
proceed with such a survey, considering the length of time that has elapsed since
the date of the seizure of the forests by Bulgaria, and considering the fact that
considerable cutting may have been done in that interval for the benefit of other
persons.
319. See Guarachi America, Inc. v. Plurinational State of Bolivia, Award, PCA Case
No. 2011-17 (Jan. 31, 2013).
320. See id. ¶¶ 452–617.
321. See, e.g., id. ¶¶ 505–510.
322. See id. ¶¶ 571–581 (noting that the sources cited by Bolivia’s expert made clear
that the expert’s “multiplier” approach, which would have increased the country risk associ-
ated with the investment at issue based on broad macroeconomic risks in emerging markets
such as Bolivia, was appropriate only in short-term valuations).
and larger macroeconomic climate.323 The tribunal in CMS v. Argentina,

for example, analyzed the key assumptions of the claimant’s valuation ex-
pert, making a number of modifications to the valuation model.324 Of par-
ticular importance, the tribunal reasoned that in determining future cash
flows from the investor’s natural gas transmission business the expert had
erred in assuming that Argentina’s economic crisis would not have re-
duced demand for gas.325 Similarly, in valuing the claimant’s investment
in an oil production project, the tribunal in Venezuela Holdings v. Vene-
zuela relied on the claimant’s oil price forecast because it included all rele-
vant information available at the time of the valuation date, including a
decision by the Organization of the Petroleum Exporting Countries not to
increase production, which would have resulted in an increase in supply
and likely decrease in price.326
Although a reliability analysis of expert evidence must to some degree
include an examination of the constituent parts of an expert’s opinion, a
legal decision-maker should in the end consider an expert’s presentation
as a whole and the totality of the evidence, and determine whether the
“weight of the evidence” supports a conclusion that the expert evidence is
reliable.327 Under this approach, a decision maker should not discard ex-
pert evidence merely because some aspects are potentially flawed but
should examine the evidence as a whole, taking into account any flaws in
that broader context. In CMS v. Argentina, for example, the tribunal
noted that the “general approach” under the DCF method of the claim-
ant’s valuation expert “remain[ed] . . . valid,” even though the tribunal the
tribunal expressed a number of “reservations” with the expert’s applica-
tion of the approach and made “a number of changes” to the expert’s
assumptions.328
CONCLUSION
Learned Hand wrote in 1901 that “[n]o one will deny that the law
should in some way effectively use expert knowledge wherever it will aid
in settling disputes,” noting that “[t]he only question is as to how it can do
323. See MARBOE, supra note 64, at 219-28; see also CME Czech Republic B.V. v.
Czech Republic, Final Award, ¶¶ 554-566, (UNCITRAL Arb. Proc. 2003), http://
www.italaw.com/sites/default/files/case-documents/ita0180.pdf (examining the factors influ-
encing an expert’s valuation and making adjustments where “certain deductions” were
“unsustainable”).
324. See CMS Gas Transmission Company v. Argentine Republic, ICSID Case No.
ARB/01/8, Award, ¶¶ 435-463 (May 12, 2005).
325. See id. ¶ 444.
326. See Venezuela Holdings B.V. v. Bolivarian Republic of Venezuela, ICSID Case
No. ARB/07/27, Award, ¶ 327 (2015).
327. See MCGARITY & WAGNER, supra note 225, at 154–56; see also BROWER &
BRUESCHKE, supra note 90, at 201 (noting that arbitral tribunals “must determine what
weight, if any, based upon the totality of the evidence, to give to the expert’s conclusions”).
328. See CMS Gas Transmission Company, ICSID Case No. ARB/01/8, Award, ¶ 434.
so best.”329 More than a century later, that question still remains, and
international law continues to evolve to more effectively use expert evi-
dence. The question also grows more complex by the day, as human activ-
ity, the disputes that arise from that activity, and the international
agreements that set out rules for addressing those disputes, also grow in
scope and their own complexity.
This Article attempts to complement existing efforts to better analyze
expert evidence, with a focus on the reliability of an expert’s methodology
and its application to a particular set of facts. A more robust, comprehen-
sive, and transparent framework built on best practices developed within
international law may assist in reaching more analytically sound decisions
and enhance legitimacy.
329. Learned Hand, Historical and Practical Considerations Regarding Expert Testi-
mony, 15 HARVARD L. REV. 40, 40 (1901).

Reliability of Expert Evidence in International Disputes

Uploaded by

Copyright:

Available Formats

Reliability of Expert Evidence in International Disputes

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Reliability of Expert Evidence in International Disputes

Uploaded by

Copyright:

Available Formats

Michigan Journal of International Law

Reliability of Expert Evidence in International

Follow this and additional works at: http://repository.law.umich.edu/mjil

RELIABILITY OF EXPERT EVIDENCE IN

* Matthew Swinehart is a legal advisor for trade and investment agreement

3. CAROLINE E. FOSTER, SCIENCE AND THE PRECAUTIONARY PRINCIPLE IN INTERNA-

stand a proposition to be justified in believing it,” they do need to “be able

I. HISTORICAL TRENDS IN THE USE OF EXPERT EVIDENCE

6. Arbitral or judicial settlement of international disputes has been described as “ir-

expropriated by the Germans, including the “sum required . . . to construct

only one cited as providing “many examples of the use of independent

were relatively few examples of reliance on expert evidence—some re-

creasingly open to nongovernmental groups, international organizations,

courage countries to adopt international standards or establish scientific

Together, these developments have significantly increased the role

38. See, e.g., notes 15-20 and accompanying text.

often has wide-ranging and significant influence on decision makers and

the legal concept of ‘necessity’ or ‘reasonableness’ in the context of the provision

Even when a tribunal or court is reluctant to rely on highly technical

II. PREVAILING PRACTICES FOR EVALUATING EXPERT EVIDENCE

Existing practices do not provide a sufficient framework to evaluate

Other efforts to improve the handling of expert evidence have focused

at hand, or to cast doubt on an expert’s testimony through cursory analysis

III. INADEQUACY OF PREVAILING PRACTICES

[T]he costs of such a tribunal-appointed expert may be considerable. Further-

(internal citations omitted).

A. Poorly Reasoned Reliance on Expert Evidence

The construction of scientific knowledge is a social process that is culturally depen-

status of evidence to “irrefutable truth.”112 Failure to assess the reliability

its complexity and legal decision makers’ lack of expertise in financial

B. Poorly Reasoned Dismissal of Expert Evidence

lenged Canadian regulations that required commercial fishing companies

tered on Argentina’s claim that a Uruguayan plan to build pulp mills on

C. Use of Expert Evidence as a Veneer to Enhance Legitimacy

reasoned, “[a]ccording to a quite general rule of procedure, the Court is

[T]he [EC — Hormones] Panel was attempting an impossible task—making deter-

of expert evidence can be so strong that decision-makers will not rely on it

IV. MORE ROBUST ASSESSMENTS OF EXPERT RELIABILITY

A. Applying a Reliability Framework

The core of this Article’s proposal is an illustrative and flexible check-

When considering the “evidential value of any evidence produced . . .

is not a formula,178 and decision makers would of course remain free to

]B. A Proposed Checklist to Evaluate Expert Reliability

178. GAWANDE, supra note 168, at 167.

1. Will expert evidence usefully aid in answering the question

The threshold question in determining the weight to accord expert tes-

experts to answer questions outside their technical competence.192 When

tions of geography and topography” because the arguments of the parties

ers with the opportunity to appoint additional experts or to require the

3. What methodologies does the relevant technical field consider

evidence has increased with the complexity of international disputes over

underlying evidence in view of the totality of the evidence presented, with-

research.235 Because policy-relevant research often addresses narrow

235. See MCGARITY & WAGNER, supra note 225, at 39.

Independent scientists can and regularly do conduct policy-relevant research in a

parties to present evidence on the extent to which advocates have commis-

4. Which expert’s methodology is most reliable in this context?

different methodologies as gleaned from peer-reviewed studies and col-

views.255 Consensus may simply remain elusive on a time-scale useful to

of science is that it is by nature progressive, with methods designed to approach a better